History

On the first of October 1862, Südbahn-Gesellschaft (South Railway Company) opened the Zidani Most – Zagreb – Sisak railway. This was one of the first railways in what are now Slovenia and Croatia, and this route is still the main rail link between the capitals of the two countries – Ljubljana and Zagreb, with the Slovenian and Croatian sections meeting at Dobova. It carried the Orient Express service from 1919 to 1977 and it remains an extremely important railroad to this day. The route was electrified by 1969, but with different voltage standards – the Slovenian section was electrified with 3kV (the same was used on the Zagreb – Rijeka railway at the time), while the Croatian section was electrified with 25kV (used in the rest of Yugoslavia). Because of this, even though passenger train service was operated by a single operator – JŽ (Yugoslav Railways), electrical service across the border necessitated engine swaps between the 25kV Croatian and the 3kV Slovenian side.

Cross Border Rail

In addition to the electric locomotive swap (performed at Dobova station which has both 3kV and 25kV power), a border check was added in 1991 following Slovenia and Croatia’s independence from Yugoslavia. This added additional complications and delays to cross border traffic and even though Slovenian (SŽ) and Croatian Railways (HŽPP) started to run relatively frequent commuter EMU train service on their respective sides, only inter city trains would cross the border. After 30 years of very limited cross-border service, border controls were abolished on January 1st 2023 when Croatia finally joined the Schengen Area. The lack of border control opens up additional opportunities for smoother and faster cross border service. Many inter city trains that operate on this line use multi-system ÖBB locomotives that can run on both 3kV and 25kV power. They often pull trains through Austria and Slovenia until they reach Dobova, where a HŽPP locomotive takes over. Without border control, using such locomotives to Zagreb (and beyond!) could shorten cross-border travel by at least 20 minutes. As of 2025, this hasn’t happened yet.

However, another service improvement option opened up following the abolition of border checks. Thanks to 3kV and 25kV power at Dobova, Croatian and Slovenian EMU trains could meet at this station and allow passengers to transfer between SŽ and HŽPP commuter trains. While this wouldn’t be fast travel, it would significantly improve travel options between smaller towns on either side of the border. Additionally, it would provide additional options for intercity travel between Zagreb and Ljubljana. In early 2025 there are only two pairs of direct trains daily, down from regular 3 due to work at the Tauern Tunnel in Austria. For example, passengers could take a train on Zagreb’s suburban line (normally terminating at the spur line station Harmica or at Savski Marof), which run at 30 minute intervals for most of the day. With the line terminating at Dobova instead of Harmica, they could then easily transfer to a SŽ train to Krško, Maribor or Ljubljana! This is pretty common in many places, with passengers connecting between different railway operators. HŽPP already performs a similar connection with the Hungarian operator MÁV at Beli Manastir station, where MÁV runs their trains to the Croatian border station, allowing transfer between the two operators.

Plans

As of 2025 there is unfortunately no such option at Dobova. There are currently 5 pairs of passenger trains crossing the border daily. The only alternative option for transferring between HŽPP and SŽ networks is to take their respective commuter trains to Dobova or Harmica and then walk/bike/drive the 2.8km between them. However, in late 2024 things slowly started moving forward. HŽPP and SŽ leadership met in Zagreb, and announced that they are looking into improving connectivity between Ljubljana and Zagreb. While this announcement did not include details about these improvements, SŽ announced only a few days later that they will be providing connections to HŽPP commuter trains in Dobova through the month of December. On the weekends, one pair of SŽ and HŽPP trains would meet in Dobova to provide connections for tourists traveling to Zagreb holiday markets. This news wasn’t widely reported — the special service was targeted at Slovenian passengers and HŽPP did not advertise it at all. But this relatively low key holiday train service provided a perfect opportunity to test out how a regular SŽ – HŽPP connection at Dobova could work in practice. It also coincided perfectly with my already scheduled visit to Zagreb, my hometown. I had already planned a train trip that would take me through Dobova on my way to Zagreb from Rotterdam (you can read more about this overnight train trip on my Mastodon). Once I got to Zagreb, I would have an opportunity to try out this cross border service myself!

The experiment

My first challenge was to figure out how to test this connection. The timetable was built to accommodate tourists traveling from Ljubljana to Zagreb. They would leave Ljubljana around 11am, transfer to another SŽ train at Zidani Most and finally transfer to a HŽPP train at Dobova, arriving in Zagreb at 1:27pm. The return trip from Zagreb would start at 7:13pm, with a single transfer to a SŽ train in Dobova that would get them to Ljubljana by 10pm. Since I was staying in Zagreb, my options were limited. I could theoretically take an early IC train to Ljubljana, and then immediately get on a SŽ train back towards Zidani most. As much as I love spending time on trains, this seemed a bit stressful. Instead I planned a different kind of trip — I would take a regular HŽPP suburban service to Harmica and then bike over to Brežice, where I would get on a SŽ train to Dobova, before transferring to a HŽPP train. While this wouldn’t be a full scale trip, it would let me observe and experience that critical connection at Dobova.

I scheduled this trip for Sunday, 29th of December, the last day of this holiday service. I took my trusty Rog Senior, made in Ljubljana. This bike was originally purchased by my father in the 1980s and I wrote about it a few years ago in my post about a BikeMS ride between Zagreb and Sisak. I got on the 10:47am train to Harmica, giving me ample time to reach Brežice and get on a SŽ train to Dobova by 12:45pm. Zagreb suburban service is primarily served by Končar 6112 EMUs which have almost fully replaced the old, Hungarian made HŽ 6111 series. These trains are not particularly comfortable for longer journeys, but they are perfect for short local trips like this one. They’ve got ample standing room and good bike accommodation.

The ride to Harmica took about 30 minutes. The Harmica spur is interesting in itself. It’s the start of the Kumrovec railroad, a single track, non-electrified 38km long line that connects the Zidani most – Zagreb main line to the Slovenian railroad network at Imeno. It crosses over between Croatia and Slovenia at multiple locations, following the Sutla river. Because of this, service was discontinued in the year 2000. Thanks to the abolition of border control, there are currently efforts to restore service. The railroad was electrified to Harmica, the first stop on this branch railroad, serving as the western terminus of the Zagreb suburban line (the east side service terminates at Dugo Selo). After a comfortable ride, I got off at Harmica and looked around a bit. Harmica is a small sleepy town. The stop is very simple — there is no station building, just a simple shelter and a medium sized parking lot. If you follow the tracks past the station, you’ll soon reach the end of electrification and suddenly the tracks become overgrown with grass, having been out of service for 25 years.

Another interesting structure is the abandoned border crossing. Even though it’s only been a couple of years since border control was abolished (and due to Schengen suspensions people may still be “randomly” checked), the structure already looks like it belongs to another era. Even though it’s a minor detail compared to real horrors that followed the disintegration of Yugoslavia, border control between Slovenia and Croatia used to be a constant reminder of missed opportunities in this part of the world. While the rest of Europe was rapidly integrating and tearing down borders in the 1990s (no Schengen suspensions back then!), in the Balkans we were putting them up. The history of border control between Slovenia and Croatia is quite interesting, and I wrote a short thread about it when Croatia joined the Schengen Zone. Even though crossing the Sutla river through the abandoned border crossing is a serene and calm experience on this sunny December morning, even the presence of this structure makes me uneasy. The existence of borders implies violence even when that violence isn’t directly present. Unfortunately, in 2025 the lack of border control feels like temporary respite.

I ride my bike towards Dobova and pass a small village called Rigonce. There’s a memorial to the Ten-Day War, the brief armed conflict that preceded Slovenia’s independence from Yugoslavia. The road isn’t very busy, and I get into town after less than 10 minutes of riding. I soon find excellent bike infrastructure, with bike paths clearly marked on the road and a fully separated path as soon as I get out of town. The paved bike path follows the road, but it has its own bridges over small streams, and it is pure pleasure to ride on it. There are picturesque fields to the south, as well as hills of the Žumberak/Gorjanci Mountains. I reach Brežice way ahead of my 12:45 train to Dobova, so I decide to ride some more. I consider continuing all the way to Krško, but I wisely decide against it. I would be cutting it close. Instead I ride around local hills and then head back to Brežice station.

Brežice station is small, but tidy. While I wait, a long freight train passes. Even though I already bought my ticket to Dobova online, I decide to use the ticket machine and I buy a paper ticket. It’s nice to have something physical to remember this short trip. At 12:45 sharp the SŽ train to Dobova arrives. It’s SŽ Series 312, a Siemens Desiro. It’s an older EMU, but it has decent dedicated space for bikes. I was hoping to get on one of SŽ’s new Series 510 Stadler FLIRTs. These EMUs are multi-system, so they could run all the way to Zagreb (SŽ ran them between Trieste and Rijeka last summer). Nevertheless, the Desiro is comfortable enough for the short 5 minute ride back to Dobova.

Before you know it, the train rolls into Dobova station. The ride is so short that nobody checks my ticket. As expected, the train is full of families going to Zagreb for the holiday markets. As we get off the train at Dobova, the PA announces that the train to Zagreb is on track 4. I follow the crowd to the underpass. Luckily, there is a simple ramp that I use to push my bike down the stairs. I’m not sure if anyone else taking this train knows how unusual it is to see a HŽPP EMU at this station. The kids are excited for the holiday market and the trains, but most people seem nonchalant about the whole affair. Meanwhile, I’m taking as many photos as I can. I think about how significant this could be – of all the opportunities that would open up if this special train became routine practice at this station. I think about employment and education opportunities, about environmental improvements and people leaving their cars at home (or maybe even selling them!). There is so much possibility hidden behind this simple train transfer.

We get on the HŽPP train and in a few minutes we head out towards Zagreb! The ride is so simple and uneventful that I need to remind myself that this is, in fact, unusual. As is often the case with borders, there is no clear indication when you cross one. I snap a bunch of photos to look back at and identify when this train crossed from the unusual into regular part of its service. I do get my ticket checked on this train. I purchased a regular HŽPP ticket from Harmica to Zagreb Glavni kolodvor online, since neither SŽ nor HŽPP sell tickets to locations on each other’s network. If you’re traveling internationally, you either have to go to a staffed ticket booth, or you can buy international train tickets at obb.at. It’s only somewhat curious that even in the 21st century the Austrian railway company gets to sell you tickets between Slovenia and Croatia, while their national operators can’t. The conductor is somewhat surprised that I have a HŽPP ticket, but she doesn’t give me any trouble. As we approach the first stops on the Croatian side, I’m trying to figure out how HŽPP arranged this trip in their timetable. Since their website doesn’t show destinations in Slovenia, there is no obvious way to find this train. Per the SŽ timetable, this is HŽPP train 8041 which normally starts at Savski Marof. While I’m still thinking about this, I decide to get off at Brdovec station. It’s still early, and it’s a beautiful day. I decide I want to ride my bike some more.

I get off and I finally figure it out! HŽPP didn’t really run 8041 from Dobova. Or they did, depending on how you look at it. Once our train got to Savski Marof, they hooked us up to another 6112 EMU, the real 8041. As far as passengers (and the online booking system!) are concerned, this is just a long train on this regular Zagreb suburban service. I chuckled at finally figuring this out and got back on my bike. There’s a new bike path that follows the main line from Brdovec to the Slovenian border. It’s paved and fully separated, but unfortunately it doesn’t directly connect to the excellent bike paths starting in Dobova. Instead, bike riders have to go around through Harmica and Rigonce to get across the small Sutla river. I ride across the border a few more times that day, before finally getting on the train back to Zagreb. This cross border rail experiment was a success!

Conclusion

Writing about this experience, I ask myself whether I’m exaggerating the potential of this cross border train service? Could this special holiday train actually be the beginning of something real? Despite decent new rolling stock, rail travel has been continuously deteriorating in Croatia for decades. Rail connectivity with former Yugoslav states to the east is non-existent. Sarajevo has been disconnected from Zagreb for almost a decade, and Zagreb – Belgrade train was cancelled in early pandemic times and it never returned. Traveling between the two cities now is reminiscent of my bike ride between Harmica and Dobova, but worse. The only way to make a rail connection between Belgrade and Zagreb is to walk 7km between Tovarnik and Šid. And could running HŽPP trains to Dobova work in these times of Schengen suspensions? This whole experiment is based on the freedom that comes from not having to check documents, but in 2025 police harass and racially profile passengers on trains all over the EU. There’s many obstacles, and I’m hesitant to hope too early. But I did actually ride a HŽPP commuter train from Dobova. It was simple, it was easy, it was comfortable and it made so much sense. I choose to believe that better times will come, and that one day we will have more trains and less borders.

The post Trainhopping at Dobova – cross border rail on a 160 year old train line appeared first on Vlado Vince.

Photography is arguably my oldest hobby. I’ve loved taking photos for as long as I can remember. I reflected on my first encounter with digital photography in 2003 in a recent post, but I’m old enough that I first got to know photography through film. I still don’t know what it is exactly that draws me to still images, but I find the same joy capturing light with a camera as I did 30 years ago.

History

I first got serious about photography when I started high school. I got my first SLR camera, a Canon EOS 3000v, in 2006. Over the next few years I used it to shoot countless rolls of film. Many were the then-ubiquitous and cheap types like Kodak Gold, which I would process at a local photo studio. Even though I was shooting film, the digital writing was already on the wall, so I would have the studio scan the photos for me and burn them on a CD. I still have dozens of disks with scanned images from those days.

In 2008 I took a course on developing photos in a darkroom. This opened a whole new world for me, and from this point on I developed a lasting love for personally controlling the whole process, from selecting the camera and film, taking the photo, to developing the film and then making a print. Soon after I built my own darkroom at home, using an enlarger from Czechoslovakia that I found on the street. At this time my preferred film was Efke KB100. I still love the classic looking style of this black and white film, but back then I used it for a very simple reason – Efke film was produced in a small town called Samobor, just outside of Zagreb, and it was widely available for cheap. Fotokemika, the company that made Efke film, shut down a few years later. I wrote a brief reflection on my experience shooting their film when I came across some long forgotten Efke infrared stock.

However, even though I could now develop my own film and make prints, I didn’t have a film scanner. Occasionally I would use a flatbed scanner to scan prints, but a majority of my self-developed film remained undigitized. Later in college, I had access to a professional darkroom and an Epson V500 flatbed photo scanner, but I didn’t like it. This scanner got the job done, but it was painfully slow, its software was clunky and the image quality was only passable. After I graduated and moved to NYC I didn’t develop any film until I discovered Gowanus Darkroom, a community darkroom that I still use often. These days I mostly develop my black and white film at home (and I’ve just started to develop color with a C-41 development kit!) and I make prints in the community darkroom. But up until very recently, a dependable scanning workflow eluded me.

Macro scanning

In 2021 I built a macro film scanning set up, using my FujiFilm XT-30 combined with macro extension tubes, a tripod, a light panel and a Lomography DigitaLIZA Scanning Mask. I built this rig after reading countless articles praising it for its speed and quality. I also purchased the Negative Lab Pro add-on for Adobe Lightroom, to assist with digital processing. My first results were encouraging. The speed of “scanning” was arguably much faster than any scanner I used before. However, with “scans” stuck on an SD card, I would have to either scan 36 images first, then review them, and then scan them again if something went wrong. While I had some early success, once I tried to replicate it, I found that there was an enormous inconsistency in my results. Sometimes I would get everything right. More often, I would deal with lens reflections, or curved edges on film, or something else. Scanning turned into a whole secondary photographic experience, since I was really taking photos of the negatives.

I occasionally deployed my macro scanning rig, but I felt like each time I used it the results were less consistent. It would take me at least 20 minutes to start scanning, and sometimes I would spend over 2 hours just to get through a roll of film. The process was unpleasant and it produced mediocre results. Of course, I could always pay someone else to scan the film for me, or I could go to the Gowanus Darkroom and use their Epson V700 or Kodak Pakon 135+ scanner. But I kept thinking that there had to be a better way to do this at home. 30+ years into the age of digital photography, how was it possible that digitizing film was still so hard?

The answer was right there. It made sense that scanning film still sucks. It’s because we stopped developing professional technology to do it.

It’s not unlike tape decks and CD players — older technologies currently going through a minor renaissance. They reached their technological peak in the early 2000s and they’ve only gone down since then. Companies trying to produce them in 2024 are stuck with bulky and low quality mechanisms because low quality mechanisms are all that can be produced in factories that long ago got rid of production lines that could make anything better. I dislike the Epson V series flatbeds, but that’s all you can get new these days, even though the models have hardly changed in decades. I realized that if I was going to find something better, I would have to look in the past.

Vintage computing to the rescue

Usurprisingly, I’m not the only person looking back to build my film scanning workflow. Decades old professional film scanners like the aforementioned Pakon 135+ may run you literal thousands of dollars. Even the more consumer grade Nikon Coolscan series scanners go for hundreds of dollars on eBay, and this is despite serious compatibility issues with modern operating systems. Luckily, this is where I had the advantage. Countless Reddit posts ask the same question — can you run Nikon Coolscan model xyz on Windows 11? What about FireWire compatibility? A few brave ones even dabble in SCSI. How about running Nikon Scan past OSX Snow Leopard? The further you go with old scanners, the more difficult they are to get working on modern hardware. Thankfully, over the years I’ve assembled a sizable collection of older computers, and I realized that building a bleeding-edge film scanning setup circa 1998 could be just the way to go.

I decided to go for a Nikon LS-2000. This is an old scanner – it was first sold in 1998 for over $2000. It can scan up to 2,700 dpi and it saves images in tiff. But most importantly, it uses SCSI, which makes it incredibly inconvenient to use in 2024. Unless you have a reasonably fast Mac OS System 9 computer that is already set up with networked storage because you recently participated in building a global AppleTalk network. Then it stops being inconvenient, and instead it begins to look quite attractive because it’s by far the cheapest film scanner on the used market today.

I started to watch eBay. I made sure to look at items that included both film strip and slide adapters. I included parts only and not-working listings as well — with devices like this, it is very rare that someone actually tests them before selling. Before long, I won an auction! I got a barely 26 year old scanner for $56 + shipping.

I decided to use it with my PowerBook G3 Wallstreet. It supports SCSI and 10 megabit Ethernet. It also has VGA video output, so I can set it up on my desk and output its video to a small screen. That way I can use the PowerBook to scan the film and immediately transfer files to networked storage, which I can access with my new M4 Mac Mini and then post process in Lightroom. I also had to get a HDI30-DB25F adapter for my PowerBook since it doesn’t have a full sized SCSI port, which set me back $20 for a new old stock adapter.

The scanner itself needed some minor work to get going. I had to disassemble the stepper motor and apply new grease, otherwise its shaft was stuck solid. This video was very helpful when doing this maintenance.

I tried using VueScan, but despite its advanced features I found it clunky and unintutive. After testing it, I went back to using Nikon Scan 3.1 for now. This program is simple, but it works well. It’s easy to set up for continuous scanning where it scans multiple images from film and saves them to an AppleShare mounted drive, following a naming convention. With highest resolution settings, LS-2000 will generate tiffs up to 57MB in size and the only bottleneck is the Powerbook G3’s 10BASE-T Ethernet. I might try to get a Fast Ethernet PCMCIA adapter to improve this. In any case, when using continuous scanning mode, the scanner averages about 2.5 minutes to scan and transfer an image to networked storage. After the first one is transferred, it roughly takes me that same amount of time to import the image into Lightroom and to do initial processing. It’s not incredibly fast, but it works and it is consistent.

Edit: reconsidering VueScan

After publishing this article, I did some more experimentation with VueScan and found its infrared clean function very effective. While I have only had minimal dust issues with my black and white negatives, the color negatives of the Ektar 100 film I recently developed myself were full of white dust on film. Some of this is due to my carelessness when washing and drying this film. However, it mean that I had to significantly edit the photos in Lightroom to get them even somewhat presentable. Setting the “Infrared clean” option in Vuescan 7.6.64 to heavy made a huge difference, as you can see in this example. I am reconsidering whether to move to VueScan as my primary software. It does take longer to scan at the same resolution, and I haven’t been able to set it to scan directly to networked drives. But the quality of its infrared clean is impressive.

This new scanning workflow isn’t perfect, but it’s probably the best I’ve had in years! My goal with this scanner was to test whether using an older, but high quality film scanner was a viable alternative to the modern “DSLR macro scanning” technique, and for me it definitely is! It’s important to consider that this is only because I already had a ready to go vintage Apple infrastructure with a PowerBook running classic MacOS and SCSI support, as well as configured networking with Netatalk2 AppleShare ready to go and integrated with same networked shares my modern computers use. Ultimately, the reason why this works for me is that I already enjoy tinkering with old technology, which is why I had a ready to go system that paired perfectly with this 26 year old scanner. I don’t think this is an easy to replicate solution for everyone, but if you already like to play around with old Macs, it might be a budget friendly solution to get good quality film scans.

Lastly — some problems. Despite the increasing hostility of modern tech companies towards their customers, we don’t appreciate enough how stable and reliable our personal computers actually are these days. In the days of System 9 computers would lock up and crash all the time, and this does happen to me often enough to be annoying. Sometimes the Nikon application will lock up. Other times MacOS will crash. It’s par for the course when working with old computers.

Conclusion

At the end of the day, despite the fun and frustrations I have tinkering and troubleshooting old technology, the reason why I embarked on this journey was to digitize film photos and share them with folks online. And since I’ve been using this new workflow, I’ve been having more fun with my photography than I’ve had in a long time! Below are a handful of photos I’ve taken and scanned with my Nikon LS-2000 recently. Enjoy!

The post My brand new digitizing workflow using a 25 year old film scanner appeared first on Vlado Vince.

Over the next few years, I got a few more portable audio players, each with more capacity, but none with the style and convenience of an Apple iPod. In 2005 I got a 5GB Olympus M:Robe MR-100. This stylish device was neat, it had the look of an off brand iPod mini but it didn’t look cheap (it also wasn’t cheap). Unfortunately, the smooth glass surface was predictably sensitive to drops and scratches, and its tiny hard drive died after one too many falls. I replaced it with a Creative NOMAD Jukebox Zen Xtra (30GB). This was an ugly, brick like device that used a regular 2.5″ hard disk. The build quality was cheap, but it was less prone to damage. Looking back, it seems like every consumer electronics company was making MP3 players at the time, and most of them were similarly unremarkable. But this device didn’t last long either – by 2008 I finally replaced it with a 120GB iPod Classic, a music player that I ended up using for a long time, well into the streaming era.

At some point though, after I was already using my iPod Classic, my friend gave me his third gen model. Its hard drive had (expectedly) died as well, and he knew I liked to tinker with things. I held onto it.

In 2011 I discovered Mark Hoekstra’s blog Geek Techique. Mark sadly passed away in 2008, but his family and friends keep his website online to this day (as of November 2024 you may get a certificate error, but the website is still up!). Mark was a tinkerer, and his blog is full of fun and interesting projects, many of which anticipated the retro storage conversions from hard drive storage to flash solutions that are so common these days (like the blueSCSI). One of these projects was the manually wired Compact Flash card adapter, meant to be used with an iPod instead of a hard drive. Mark may have been the first person to do this – it seems that such adapters weren’t a mainstream thing before then. While there are references to him trying to commercialize this project, he sadly passed away soon after. Cheap CF to 1.8″ IDE adapters became available to buy online by third parties before long. I still visit Geek Technique from time to time, and I’m happy that Mark’s work lives on, on line. I wrote about his blog earlier this year on Mastodon, and a few years ago on twitter (archived post).

Prompted by this discovery, I decided to attempt to repair my friend’s third gen iPod. I got one of those cheap adapters on Amazon, but I never got it working. Frustrated, I replaced the 1.8″ drive with a used one. I seldomly used this iPod over the next decade, and the replacement hard drive still works. In 2020 I briefly attempted to get the Compact Flash adapter working yet again, while I was restoring my black Macbook (2006). This attempt also failed, but I did document using the third gen iPod to natively download podcasts through iTunes. As of 2024 this is almost impossible, because virtually every RSS host requires modern SSL, which is no longer supported by old versions of iTunes that did support iPods and podcasts.

Fast forward to earlier this year. Prompted by the resurgence of interest in pre-streaming music players like iPods and Walkmans, I took out the old third gen iPod once again. I scoured the internet once more, looking for documentation on successful flash conversions. I found the testimonials lacking — almost none went into technical detail, and they primarily dealt with setting up compatible versions of iTunes to successfully restore and format the card with iPod firmware. I already had a good version of iTunes — I used it without issue to manage the hard drive in my iPod.

I was using the same CF to 1.8″ adapter I got in 2011. I tested continuity countless times, and I couldn’t find anything electronically wrong with it. As soon as I put it in, the iPod wouldn’t start at all, stuck on the Apple logo. I tried tons of CF cards as well as SD cards in a CF adapter. I could not even get to the point of attempting a firmware restore. I attempted using different iTunes versions and computers to try to get it detected. Nothing worked.

Then I decided to try one thing that I hadn’t done before – I got an actual FireWire to 30 pin cable, which is needed to restore an iPod 3rd gen from a Mac (Windows can do it from USB). I used OSX Tiger because later iTunes can’t use FireWire. I had no luck at all.

Finally, I actually managed to get some sort of recognition, but not in the way that was actually useful. Through bizarre trial and error, I discovered that if I format the SD card inside a CF adapter to an HFS+ volume, then don’t put it in yet, but turn the iPod on via battery, and wait for the folder icon to appear. I would then connect FireWire, followed by a quick insert of the CF adapter and then the iPod would actually switch to disk mode and actually get recognized by my Mac!

But this was still not working. The Mac would mostly freeze, and the few attempts at restoring failed. I finally gave up on the adapter I bought 12 years ago, and I ordered a new one for $7. It still made no sense to me why the adapter would be the cause of the issue, but I had to try something else.

I got the new adapter. It was the exact same design, with identical connections and I had no idea why it would work any better. However, unlike the first adapter, this one actually showed me the folder icon immediately without my weird trick!

Attempts at restore failed, again. But I kept going. I did have an image of a properly formatted drive, so I copied that over to the SD card with dd.

$ dd if=ipod.img of=/dev/sdb status=progress

I had actually tried this earlier, so expectations were low. Except this time, the iPod happily booted! However, dd’ing a 20GB hard drive to a 64GB SD card left lots of empty space.

Since I couldn’t use iTunes to restore, I backed up the contents of the storage partition, then I used gparted to delete it and create a new one, this time extending it to full size of the 64GB SD card. Finally, I copied the files back over. The iPod and iTunes didn’t mind. This was the partition table after this was all done:

Once this was done, I was able to sync the iPod with no issues over USB. This decade long repair did leave me with some unanswered questions. I know that the original adapter must have had some kind of fault. Maybe there was some undetected short or something that made it impossible to start booting with it, but if you engaged it later it would function. Who knows!

The issue with restoration still bugs me. I don’t know why my attempts to restore via FireWire kept failing, even with a good adapter. It’s possible that in the end it came down to a specific version of iTunes and OSX version. My solution of just dd’ing a known good iPod 3rd gen hard drive and then recreating the storage permission was a very brute force way of getting around this problem. Paul Rickards recently did a much more elegant iPod 3rd gen restoration where he recreated those partitions manually, and then copied over iPod firmware with dd.

It’s important to note that I had different experience with different 30 pin cables. I was unable to successfully use the 3rd party FW cable, nor the Apple USB/FireWire combo Y cable. The USB in the Y cable worked sometimes, but not consistently. All my successful syncs were done using a regular 30 pin USB cable. Whether something is broken in those FireWire cables, I can’t say.

In any case, my iPod has been reliably working for months now, and I’ve been using it as my primary portable music player. I barely use my smartphone to listen to music any more. The iPod lets me listen to exactly what I want – no more random Spotify versions of songs that I know should sound different, or unintended switching into “artist radio” mode when I want to listen to an album. It also helps me stay off my phone when I’m riding the train or walking. It’s a strange and frustrating thing when you have to use older technology to keep yourself from engaging in compulsive behavior caused by contemporary tech. The only feature I really do miss is the ability to load up podcasts. The third gen iPod was one of the first portable players actually used for podcasting, first with dedicated programs like iPodderX and later with iTunes once Apple implemented the functionality, helping give podcasting the name it still holds. I could build an SSL proxy server to help with this issue, but I haven’t had the time yet.

Which lastly brings me to library management. I’ve been using my 2006 Macbook to manage music on my 2003 iPod. I like how it makes the process intentional, and I usually update music every few months. The third gen iPod actually does get recognized by macOS 15 though, and I could theoretically use my brand new Mac Mini to manage it! Apple moved iPod management into the Finder after it sunset iTunes, but my experience has been mixed. For now, my iPod is still managed by its contemporary Macbook.

Useful files:

• 3rd gen iPod 20GB hard disk image (2GB)
• Storage partition (Apple_HFS) files

The post Flash modding a third gen iPod in 2024 appeared first on Vlado Vince.

The October Archive.org attack shook me in a bad way. For a while now I’ve been increasingly convinced that in the not-too-distant-future we will be referring to the period of the early 21st century as a digital dark age. Not just because of the seemingly unstoppable rise of anti-intellectual, anti-science fascist movements around the world that use the internet to spread like wildfire, but also due to significant gaps in organized preservation of digital data, both public and personal. By this point we’ve seen enormous swaths of early internet go away on a whim of owners of previously significant platforms – from Geocities to MySpace. But while there have been organized efforts to collectively archive data when it comes to such large and well known services, I’m increasingly finding that significant parts of the web tend to go offline not with a bang but a whimper. And the speed of this loss seems to be increasing, seemingly in parallel with Google’s enshitification of search and the rise of halucinatingly unreliable “AI” tools that normal people are starting to use not just as replacement for search, but as a primary source itself.

While I’m primarily concerned about this phenomenon in terms of its impact on public data – information that in previous decades would have been published on paper, records, tapes or other physical media and then distributed publicly, I’m also increasingly worried about personal data. Let’s take photos for example — while folks would lose physical photographs all the time in the pre-digital times, from simply losing them in a move to experiencing catastrophe and trauma in war or a house fire, the physicality of those memories largely ensured that at least certain important pieces would get preserved through decades, and sometimes centuries. When my grandfather passed away three years ago, he left photographic evidence of his life spanning 90 years. On the other hand, I already have a digital photographic gap in my personal history. My family got our first digital camera some time around 2002. Due to a hard drive failure on my dad’s laptop, which occurred some time into that first year of digital photography, we lost a whopping 2268 digital photos taken by the Fujifilm FinePix S602 Zoom, our first digital camera. We still have lots of photos from the 90s, but there’s a noticable gap in the early 2000s, meaning that I have very few photos of me around the age of 11 and 12.

Personal history

How do I know that we lost over two thousand photos? Because the earliest digital photo I have in my personal archive is dated March 15, 2003, and it is titled DSCF2269.JPG. It was taken at CeBIT in Hannover, a technology fair my father took me to in 2003 and 2004. It shows two AIBO robot dogs, with a crowd gathered around them. I think it’s quite fitting that this is the first digital photo I have preserved, a great example of the early 2000s techno utopianism. I was 12 years old, and I still remember the excitement. I’m sure that after March 2003 we lost more digital photos, but at that point I spontaneously started preserving digital photos and videos, first those taken by our shared family digital camera, and later those taken by my own cameras and phones. I often wonder how other folks have fared in this effort. While backups and preservation have probably gotten more consistent with services like iCloud and Google Drive, outsourcing this effort to the same companies engaged in hyper enshitification and “AI” slop worries me. I’ve helped a number of friends recover photos from old phones, and I know that many of them have huge gaps in data from the 2000s. I can’t even speculate what’s going on with younger folks who grew up in the post personal computing era, some of whom seem to exist in a post file system world altogether and primarily use ephemeral social media.

But speculation and fear are not the topic of this post. Instead, I would like to document my own personal history preservation effort which I executed over the last 12 months, putting in many hours into trying to reconstruct years 2007 – 2012 of my own blog, this same blog that you are reading right now. But first, we need to establish some general history.

Early blogging era

I started my first blog in 2006. I was in high school, and I was obsessed with photography. I already had a cell phone with a camera (Nokia 3650), but I had recently gotten my first digital camera: a FujiFilm A350. I was taking tons of photos, and at some point I discovered the joys of street photography. I was quite impolite and lacked any subtlety. I was taking photos of people young and old all over Zagreb, my hometown. Blogs were hot at the time, so some time in the Fall of 2006 I started posting my photos on a local blogging service – blogger.hr. My blog was hosted at “onthe.bloger.hr”. There was no particular reason for that name, naturally. I became a very active photo blogger — I posted on there all the time. This being the “middle internet” era, I quickly made friends and started meeting up with the local photo blogging community. I was the youngest person in this group, but there didn’t seem to be anything odd to me about being a 15 year old hanging out with a bunch of 20 and 30 somethings. We were all learning how to take good photos and having a good time together. This wasn’t my first website — I had dabbled in free website services earlier, but this was the first time I maintained a consistent and continuous presence online that was actually connected to me as a person in the real world. Those were fun times — after a few months, I got featured in the free Metro newspaper. Remember when they wrote about cool websites in the newspapers?

In 2007 I decided to go independent – I left the local Croatian blogging service and I set up a WordPress blog with my own domain – mejs.org. Everything old is new again, and almost 20 years later this blog once again runs on WordPress. “mejs” is a handle I had been using since the early 2000s. I was a prolific poster on Star Wars and other adjacent science fiction message boards, so I came up with a localized moniker based on Samuel L. Jackson’s character Mace Windu. “Mejs” is the phonetic spelling of the English word “mace” in Croatian. Wasn’t the old internet great? I carried this handle from forums through blogs and Twitter, and I still use it on various services including Mastodon. I lost access to the mejs.org domain sometime in the 2010s, but I regretted it immediately. I kept an eye on it, and once the domain squatter who held it for the better part of last decade failed to renew it, I went ahead and grabbed it back. If you type in mejs.org nowadays, you’ll get to this website. But what happened to the original mejs.org?

Abandonment and regrets

I launched mejs.org in June 2007 as a single website that would host two blogs — my “photo blog” and my “text blog”. The distinction stemmed from the fact that I struggled to represent my photography in a visually appropriate way while maintaining ability to write articles on the same website. I posted A LOT. The website went through a few transformations, but by 2012 when I was already in college and living in Portland, Oregon, it was almost dead. This was for a few reasons — I was increasingly interested in different things, blogs seemed to be on their way out anyway, and last but not least I was living in a foreign country while still maintaining a website that was primarily written in a language that none of my new friends understood. I wasn’t looking back at that time — my life was changing constantly, and the same thing was happening elsewhere in the world – it seemed like something new was happening all the time. It’s funny now, looking back, how so many rigid, unchanging things that are actively harmful nowadays date to those exact years when the world seemed to be changing so quickly.

I decided to change too. I shut down my personal blog in Croatian and I opened a tumblr, redirecting the blog.vladimirvince.com domain towards it. I continued posting photos, with an occasional written post. I also used the website to present a brief biography, since professional presentation increasingly became important as I was getting closer to finishing college and diving into the life of job applications and work. Some time in this flux, I stopped paying for my hosting service, and I lost the mejs.org domain.

I did not backup the WordPress database.

Recovery

It would be years until I realized that I had no backup, and even more years until I realized that this was something I actually cared about. I slowly started doing personal digital archaeology about 5 years ago. I’ve been relatively successful in maintaining backups of my computers and servers over the years, but my web presence hadn’t been as meticulously maintained. After all, we used to be told that the internet was forever. My first attempt at restoring mejs.org was running a web crawl on the Archive.org’s backup of the website. This was surprisingly successful, and it helped establish some semblance of structure to what became a year long reconstruction effort. It wasn’t until last year, once I finally recovered the mejs.org domain, that I seriously engaged with actually recovering the full scope of my original blog, from the years 2007 and 2012 (I decided to forfeit the onthe.bloger.hr era).

I started the restoration effort while posting on Mastodon. I was browsing the Archive.org version of the blog circa 2009 and noting how good it looked. This finally prompted me to see if I could actually rebuild the whole thing. I started manually, going through the Wayback Machine archive and manually entering posts into this blog where I’m writing now. While the Wayback Machine is incredible, it’s easy to miss content published in between crawls. Sometimes there would be links that suggest a post existed on a certain date, so all I would get is a title. In other instances there’d be an archived post, but the image would be gone because it was hosted on flickr. I did have a backup of my old flickr with json files, so I could sometimes crossrefernce this to exif data and then dig out the image from my archives. This was painstaking work.

After a while I discovered an actual XML export of the database that covered content from May 2007 to March 2008! While this did not include attachments (photos being the primary content of this blog), it helped validate what I was getting from Archive.org. While the XML export was good source of truth, it is difficult to parse, so I used wordpress-export-to-markdown tool to get markdown of every post, which also pulled all media that was available online (a surprising number of photos hosted on Flickr are still available, even though they’ve been “deleted” on the account). I then ran another tool (Showdown) to convert the md files into html. Finally I went through them one by one, comparing it to my recovered archive on my current website.

But beyond the difficult restoration work, it was fascinating to see what I was up to almost 20 years ago! I posted A LOT. It was primarily photos, but I posted almost every day and sometimes multiple times a day. No wonder I embraced Twitter early. There was so much engagement in the comments! Some of it was from my friends, but a lot was from random folks who would just happen upon my blog. There were also frequent notices of travel, where I would give folks a heads up that I might not have internet access for a few days. Incredible how strange that seems now.

The XML export ended in March 2008 – just when I launched the new and improved version of mejs.org. Gone were the days of separate photo and text blogs — I was now posting everything together, on a rather standard WordPress template. I still posted a lot, but the restoration work after this point had to fully rely on the Wayback Machine. While I think I’ve managed to track down every post from this era, a decent number of them do not include photos, since I haven’t been able to identify what was originally posted.

I launched the last major revision of mejs.org in early 2009. I took yet another approach – I subtitled the blog “artistička radna organizacija”, referencing the 1980s Yugoslav New Wave compilation record Artistička radna akcija. The slightly pretentious title was meant to embrace my stated intent to expand the scope of my writing to art in general – I planned to write not just about photography, but also about film, music, books. Looking back, I certainly made an effort, writing about shows I saw or about my friends’ artistic efforts. I even engaged in a minor publicity stunt, taking the website offline for 2 weeks and then organizing a Facebook event around the re-launch of my blog. This was truly representative of me as a junior in high school, and this blog format remained the structure for mejs.org to its end in 2012.

Out of all 3 major eras of mejs.org, this one had the worst quality crawls by the Wayback Machine. Attachments were almost universally lost, and there are whole months of posts missing. I continued searching for any trace of actual backups. From my current perspective, I’m petrified at the idea that some former me just nuked years and years of writing and photography without taking a single backup between 2008 and 2012! Eventually I discovered emails with my old web hosting provider in 2011 when my wordpress was apparently hacked (I have no memory of this), and they helped me rebuild the site on another server. I found an email I sent on April 6, 2011 saying I downloaded the WordPress database. I must have used this copy to restore the blog on the new server, but despite my best efforts I never found this copy. The internet is, in fact, not forever.

One saving grace is that post 2009, my posting went down significantly. I was busy with life offline, while my posting slowly but surely moved to Twitter. This meant that even though the archive.org crawls had huge gaps, they didn’t miss all that much. My recovery effort ended up taking over a year, but by September 2024 I finally got it done. To the best of my knowledge, November 2011 is the last time I posted on mejs.org. On April 17 2012 I started posting on Tumblr using the blog.vladimirvince.com domain. Perhaps there were some posts in between those dates, but for now I’m calling this effort complete! You can find all publicly available posts in the month-by-month archive on the right hand side Archives list.

My Tumblr-as-a-website phase was relatively long. But, by September 2019 I was once again back running WordPress, and this blog hasn’t changed much in the last 5 years. Luckily, I properly migrated all the Tumblr data, so this blog hosts all the content I posted since 2012. These days I put more effort into writing than web design or formatting, and while I post only seldomly, I primarily post for myself. I am thrilled if others find my work interesting or useful, but at the end of the day, this is my web log, and I’m excited to be once again writing on it.

Conclusion

So what did I learn? Restoring my old blog was a strange experience. I’ve spent a lot of time in recent years looking back. For me, so much of looking back is thinking about the world before the pandemic, but this restoration effort primarily placed me in my late teens and, crucially, into my life before I lived in America. Reading my own words from that time, in a language I no longer speak out loud every day, is both fascinating and somewhat distressing. Some of this is, I presume, normal for a person getting close to middle age. I started this post reminiscing about photography, and the challenges of preserving personal digital archives. I never wrote a diary with any consistency, but my blog is a unique insight into my life and my mind in my formative years. Later on, Twitter would take that role for me, but I’m content with limiting the scope of the public Twitter archive I maintain. Not every thought needs to be preserved online for all to see.

Perhaps more than anything, this process made me appreciate the immense, colossal importance and value of the Internet Archive. It is as much an integral part of our common human information infrastructure as the Wikipedia, and experiencing life without the ability to browse the Wayback Machine or access some archived books for a few days was literally terrifying. Writing this days before yet another election where a fascist is threatening to dismantle what remains of American democracy, and in the middle of the ongoing genocide of Palestinians in Gaza, the stakes of information preservation are so much higher than the silly personal history restoration I wrote about here. While there are so many more actions we should all be taking to resist fascism, regularly donating to the Archive.org is one of the things I will be doing from now on.

The post Personal digital archaeology: recovering this blog 2007 – 2012 appeared first on Vlado Vince.

But that all changed this year, when the ship was evicted from its dock by Penn Warehousing, which owns Pier 82. I hadn’t followed the story about the ship’s status too closely over the years, but it seems like the writing was on the wall when their docking fees were more than doubled in 2021. By October it was clear that the plans to stop the eviction failed, and once that SS United States Conservancy signed an agreement with Okaloosa County in Florida to transfer ownership to them and sink it as an artificial reef, I started making plans to visit the ship before it’s gone. Rumors spread that the SS United States may be gone as soon as Friday, October 24th. I was skeptical – I can’t imagine that an old ship that has been sitting in dock for 28 years can just be towed hundreds of miles to Norfolk Mobile, Alabama (where it’s supposed to be stripped and prepared for sinking) on short notice. And rightly so — as of this writing on October 27th the ship is still in Philly. However, this urgency prompted me to head down and see her before she is gone.

On Sunday October 20th I got on a 7am Keystone Service Amtrak train at Penn Station, and an hour and a half later I was at the beautiful Philadelphia 30th Street Station. I rode my Brompton on the quiet South Philadelphia streets toward the Port of Philadelphia. It was a sunny, slightly chilly Fall Sunday, and the streets were almost deserted. It was a beautiful ride, and I could notice the city slowly waking up – businesses starting to open, folks taking dogs for a morning walk, churches starting Sunday service. Somehow, every time I visit Philadelphia it’s a beautiful day, and this one was no different.

I slightly overshot the turn for Dock 84, so I ended up approaching the ship from the South. Its size was breathtaking! Its two large funnels dominate the sky and can be seen from far, far away.

Once I reached the ship, I was faced by an expected obstacle. The pier is surrounded by fencing, and the owner did not facilitate any kind of viewing platform or access, not even now that the ship is departing after 28 years. This was unfortunate, but expected. There were a handful of other people who showed up this morning to say goodbye, and we traded tips on where to find tiny gaps in fencing that might be used to take better photos. A pair of somber goodbye signs and some flowers were attached to the central part of the fence.

I carried 3 cameras with me – my Fuji XT-30 with a 27mm lens for wide shots, my trusty old Canon EOS 5D with a 75-300mm telephoto lens for closeups, and my even older EOS 3000v with a 28-135mm lens and expired Efke 25 film. The telephoto lens ended up being pretty good at taking photos through the fence.

Sadly, the starboard side of the ship was barely visible through the fence, because of the mesh cover installed on that section. I couldn’t escape the feeling that Penn Warehousing actively wants to dissuade people from seeing the ship up close. It makes sense — they have obviously had a hostile relationship with the Conservancy, and want the ship gone. But it was still shocking to see how little effort there seems to be from any wider community or even governmental body to facilitate a proper goodbye to this historic American ship. When it was originally built, the US government covered two thirds of its manufacturing cost, because of their requirements for potential conversion into a troop ship. 70 years later, there isn’t even a hole in the fence or a viewing platform to say goodbye.

The pier itself was bustling with activity. Large trucks were moving materials around the dock every few minutes, and the view of the ship itself was often blocked not just by fencing, but also by large piles of pallets and containers.

In addition to the folks like me, taking photos through the fence from the street, I saw at least one drone as well as a person walking on the bow of the ship. I couldn’t tell if they were working on the ship, or if they got a highly coveted tour before she departs on her final journey.

Seeing the ship up close was absolutely worth it. I’ve seen many photos and videos of it before, but you really have to see it in person to appreciate its size and clean lines. And staring directly at its bow you can see how narrow this enormous ship is! From what I understand, it was made to be narrower than its contemporaries in order to pass through the Panama Canal, and to help with speed. Its beam is 30.9 meters, compared to RMS Queen Mary’s 36 meters.

After an hour and a half of trying to capture the ship from every angle, I crossed the wide Columbus Road and headed to IKEA for some well earned meatballs. The view of the ship from the restaurant is absolutely stunning, and Philadelphians have been blessed with a unique view of maritime history while buying ready-to-assemble furniture.

I said goodbye to the SS United States for what will likely be the last time. Though who knows? Maybe this departure will be postponed again, or maybe I’ll get to see her on the bottom of the ocean one day.

Before heading back to Brooklyn I met up with a friend whom I haven’t seen in many years. We walked the streets of Philadelphia for a few more hours. I finally got on a SEPTA train to Trenton, and then transferred to a NJ Transit train to Penn Station. While my 7am Amtrak ticket only cost me $15, the day-of tickets to NYC were 10 times more expensive, so I opted for the scenic route on local trains.

Speed is a funny thing. The SS United States was built to be the fastest ship out there, but even as she set sail at the dawn of the jet age, planes were poised to replace ocean liners as the means of travel between continents. She maintained service for 9 years, cruised for another 8 and has been virtually immobile for the last 50 years. As much as I would love to see this ship restored and preserved, I understand why it was always unlikely. Even the famous RMS Queen Mary has been struggling at its Long Beach home (I was lucky to visit it last year!) and I can imagine that its enormous costs serve as a cautionary tale for all cities that may consider adopting an aging ship.

There are some more hopeful maritime preservation stories closer to where I come from. The M/V Marina, the Jadrolinija ferry built in 1936, was preserved as a hotel in Rijeka. I have fond memories of sailing on it between Zadar and Silba as a kid. I had dinner onboard a few years ago, and from what I can tell this conversion has been a success.

Another preservation effort is currently underway in Rijeka. After decades of uncertain future, Josip Broz Tito’s famous ship Galeb is currently under restoration to be open to the public as a museum ship. Also known as the the Peace Ship Galeb, this Yugoslav Navy training vessel served as Tito’s personal yacht on many around the world trips as one of the leaders of the Non-Aligned Movement. It has also been slowly deteriorating for the last few decades, mostly in Rijeka. But unlike the SS United States, it has seen adaptive use over the years. I visited the ship twice – once in 2011 for an exhibition of Yugoslav art, and again in 2013 when it served as one of the venues for the REPUBLIKA festival. It was protected by the Croatian Ministry of Culture in 2006, and its current restoration is primarily financed by European Union grants.

In the case of the SS United States, there doesn’t seem to be comparable public interest, nor funding to preserve it. While millions have likely seen the ship from the IKEA restaurant and parking lot, I understand why the pier owner wants it gone. And if it is turned into an artificial reef in the end, it will be a better end than going to a scrapyard.

You can see more photos from my visit to the SS United States on my Flickr.

The post Saying goodbye to the SS United States appeared first on Vlado Vince.

Earlier this year I had lots of fun participating in GlobalTalk, a fun collaborative experiment where a bunch of folks interconnected their home AppleTalk networks into one global network that allowed their classic Macs and other devices to communicate across vast distances. While AppleTalk is normally tied to a physical location (contemporary uses were usually limited to offices or schools), I was eager to experiment with using it on the go.

With ubiquitous mobile internet these days, this shouldn’t be hard. However, Apple Internet Router, the software we’ve been using to route AppleTalk over the internet, relies on UDP port 387, which it needs to expose to other routers it’s connecting to. At home, that’s easy — we all just open that inbound port in our firewalls and forward it to the AIR internal IP. If you’re lucky, your ISP may provide you with a permanent public IP. If you’re not lucky, you can still set up dynamic DNS to a domain name. However, on the move it’s a different story. Normally you don’t get a static IP on your mobile connection and you can’t open up inbound ports. And if you’re trying to use someone else’s WiFi or wired internet, you still need them to make those firewall changes. Not very practical when you want to be on the move.

My solution? VPN and a bunch of small, physical equipment. Here’s the general idea:

• an OpenVPN server with a static public IP and firewall is the end of our “internal” network and the edge to the public internet. Its public IP is seen by the rest of the GlobalTalk network. It port forwards UDP 387 to its only VPN client (OpenWRT router). It runs in AWS but it could be anywhere.
• an OpenWRT router installed on a Raspberry Pi connects to the OpenVPN server via whatever internet connection is available. It can be Ethernet or WiFi via USB dongle. OpenWRT port forwards UDP 387 to AIR running in QEMU on a MBP that’s connected to the OpenWRT’s internal network
• The OpenWRT router is connected to an ethernet switch to establish its internal network
• A MBP running QEMU with AIR is setup with USB ethernet passthrough. This connection goes to the ethernet switch. AIR is setup with IP on the OpenWRT “internal” network, and the OpenWRT is forwarding UDP 387 to that IP
• AppleTalk devices are connected to the ethernet switch. Those that support AppleTalk over Ethernet connect directly. A TashTalk is connected to provide LocalTalk connections via daisy chained dongles

But where would I deploy this? While I chuckled at the idea of running this on a subway train, I decided to target VCF Festival East. GlobalTalk was such a big thing for me in the spring, and I thought it would be fun to show other folks what some of us have been doing with our networked classic Macs.

My plan pretty much worked exactly as outlined above. There were some hiccups — I initially tried using a cheap router with OpenWRT and spent hours trying to get it to run properly. Finally I moved to a Raspberry Pi as the OpenWRT router and never looked back. My primary reason for trying to use a network device was to cut down on physical equipment needed — the cheapo router had a built in switch. But using the Rpi + extra dongles for Ethernet wasn’t horrible. I still packed the cheap wireless router, in case of emergency.

Ultimately, this setup easily fit inside a backpack, but it was bulky. Because AppleTalk requires physical Ethernet (or at least it needs to think it’s using physical Ethernet), the whole configuration used lots of cables and dongles. Had I managed to emulate AIR directly on the RPi — perhaps by running Josh Perez’s jrouter — this would have been a lot simpler. But this was just around the time they were developing it, and AIR inside QEMU was the most flexible Apple Internet Router deployment I had available to me.

With a few hours to spare the night before VCF East, I successfully tested the mobile GlobalTalk setup from my LocalTalk enabled PowerBook. The connection didn’t touch my home network — it was all running over my cell phone connection!

With the setup successfully tested, I set out for VCF East. I travelled by bike and train, so this setup was truly certified “mobile”!

Once I got there, I tried to get connected by Ethernet. I had reached out to the organizers earlier, asking if they could provide a port or a cable. Unfortunately, that didn’t pan out. VCF is a guest at the InfoAge Science Center, and network infrastructure is somewhat hard to come by in the hundred year old Camp Evans buildings. Not an issue for the mobile GlobalTalk setup though — I was able to connect to their public WiFi. It was a good call to bring the wireless router with me, because the small USB WiFi dongle on the RPi couldn’t maintain a strong enough connection. I switched out the wireless router into wireless bridge mode and connected to VCF East public WiFi. We were good to go!

And this is where, sadly, this story mostly ends. I didn’t have a table this year, so I offered the connection to some other folks. They were very gracious to host me at their table, but our attempt to share their ImageWriter II over GlobalTalk unfortunately failed — due to an issue with TashTalk (which has since been fixed!), I wasn’t able to extend the LocalTalk connection to their printer. We ended up running the connection for the duration of the event, and some dear friends left us messages in the networked share I set up inside the temporary VCF East zone.

Nonetheless, I consider this experiment a success! In retrospect, there are two things I would have done differently

• beyond just having a proof of concept, I should have confirmed *what* to do with the GlobalTalk connection ahead of time. That way I could have anticipated ImageWriter II + TashTalk compatibility problem
• I should have used a real device instead of emulating QEMU for AIR. While emulation allowed for remote access, had I used my PowerBook G3 (which I had with me!) to run AIR, it would have handled LocalTalk bridging on its own, without a need for TashTalk

Things have quieted down since the peak of GlobalTalk, but folks have continued developing since then. I’ve mostly disassembled my mobile AppleTalk rig, and if I were to build it again I would probably uses Josh Perez’s jrouter. Make sure you also watch their excellent talk about GlobalTalk!

The post Taking AppleTalk on the road: a mobile GlobalTalk retrospective appeared first on Vlado Vince.

This article covers two issues I ran into:

• How to get into Recovery mode on Windows 10/11 if the regular method of Restart + Shift or interrupting boot 3 times doesn’t work
• How to get into Safe Mode even if you get blocked by BitLocker

Getting into Recovery mode via SecureBoot toggling

This particular machine that was affected by the faulty CrowdStrike driver kept BSOD-ing, but it wouldn’t actually get to Recovery mode. Instead it would just boot back to the login screen and BSOD again.

I couldn’t get into Recovery mode by holding Shift while choosing Restart during the 2-3 seconds before it BSODs, and the method of interrupting Windows login by powering off the computer 3 times didn’t work either.

I accidentally discovered that a way to get into Recovery on this BitLocker enabled machine was to go into BIOS, turn SecureBoot off, and then use the BitLocker error to get into Recovery. These are the steps:

• Disabled Secure boot in BIOS and Rebooted
• At BitLocker screen, I clicked escape to get to the prompt where I can select Startup options by clicking F1

• Rebooted
• In BIOS I re-enabled SecureBoot
• Got back to the BitLocker screen. If you have a BitLocker key, enter it now and continue into Safe Mode. If you don’t have it, follow the steps below to get into Safe Mode from regular boot
• Got to Startup options and continued into Safe Mode

Getting into Safe Mode with BitLocker enable

Like I noted above, to get into Recovery Mode by toggling SecureBoot on a BitLocker enabled machine, you do need BitLocker keys. Since I was assisting a third party, I didn’t have the BitLocker keys for this computer. I followed these steps to enabled default boot into Safe Mode that allowed me to boot without having to enter BitLocker keys:

• Follow the same steps as above to get into Recovery mode
• Go to Troubleshoot > Advanced options > Command prompt
• Get a command prompt open, skip the warning about not being able to access the local drive due to BitLocker
• Run bcdedit /set {default} safeboot minimal or  bcdedit /set {default} safeboot network to boot into Safe Mode with networking
• This forces the computer to boot into Safe Mode
• Reboot (SecureBoot should be re-enabled as above)
• Take the steps necessary to mitigate the issue (go to C:\Windows\System32\drivers\CrowdStrike, delete all C-00000291*.sys files)
• Reboot again, follow the SecureBoot BIOS steps do get back into Recovery mode
• Run bcdedit /deletevalue {default} safeboot to disable default booting into Safe Mode
• The computer will now boot normally

I hope this helps some folks! Huge props to all IT teams working on this today. I got a weird and “safe” opportunity to mitigate this on a single machine for another organization, and working with their team was a humbling insight into how hard so many IT folks are working today. They are the real heroes of the day.

The post Mitigating the CrowdStrike outage without BitLocker keys + how to get into Windows Recovery mode via SecureBoot toggling appeared first on Vlado Vince.

In the 2020s we’re celebrating many anniversaries related to the early days of home computing: 2022 marked the 40th anniversary of the ZX Spectrum and the Commodore 64, and the recent celebration of the 40 years of the Apple Macintosh made a big splash in the mainstream media. In another part of the world 40 years ago, the Galaksija computer and its accompanying publication Računari u vašoj kući arrived on the scene. They changed the lives of thousands of people in what used to be called Yugoslavia.

While the story of the Galaksija almost seemed forgotten in the 1990s and the early 2000s, this unique computer has gone through a significant renaissance in recent years, in the countries of the former Yugoslavia and beyond. If this is the first time you’re hearing about the Galaksija, I recommend you read my Making A Galaksija Computer FAQ and check out my talk at HOPE 2022 to learn more about how the Računari u vašoj kući magazine and the Galaksija computer brought microcomputers to Yugoslavia at a time when virtually no-one was able to get them legally. You may find this story fascinating and relatable even if you have no particular interest in the history of technology from formerly non-aligned socialist countries.

Many communities dedicated to famous computers like the ZX Spectrum and the Commodore 64 organized tributes to mark these computers’ anniversaries. But the Galaksija is getting celebrated in a truly special way: Dejan Ristanović and Voja Antonić, the duo behind the original project, have teamed up once again to celebrate both the Galaksija computer and the publication that made it possible, by releasing beautiful reproductions of both!

Računari u vašoj kući reprint

Whenever I write or speak about the Galaksija, I always stress the importance of the magazine that launched it. Računari u vašoj kući (“Computers in your home”) magazine didn’t just make the Galaksija computer famous – it literally made it possible. Galaksija was conceived, and largely realized, as a primarily DIY design that was freely distributed through a magazine. Računari u vašoj kući had to achieve so many simultaneous things: educate a largely uninformed public about what computers even were, explain to that same public that they themselves could construct an actual computer themselves, and then provide clear enough instructions and access to parts so that those folks could legally build a fully functional microcomputer. This was an achievement of enormous proportions.

Even though the magazine went through multiple prints, with tens of thousands copies sold at the time, the original issue is difficult to find these days. It was printed on cheap, newspaper style paper, and the quality of the print was poor even in the 1980s. Even though it’s been scanned and available online for many years (including the English translation of the instructions published in PoC——GTFO, Volume 2), the printed version is still the best way to appreciate this seminal moment in computing history.

For the 40th anniversary, Dejan Ristanović, the creator of the original Računari u vašoj kući publication, prepared a brand new reprint that painstakingly recreated the original publication from scratch, and re-published it through the PC Press magazine. As Dejan notes in the publication accompanying the re-issue of Računari (appropriately titled Računari u vašoj kući 40 years), the project of reissuing the magazine actually took over 15 years itself – it was first proposed back in the 2000s when a significant chunk of work was completed reverse engineering the typeface and design standards used in the olden, analog days. If you speak Serbian (or with a use of a translator app), the publication describing the 40th anniversary project is a treasure trove in itself – recreating a 40 year old publication with no digital files is an enormous project. Dejan and the team recreated the whole thing from scratch, including taking new photographs of Galaksija assembly and tracking down images originally published in American and British computer magazines which were reproduced back in the 1980s by simply copying them (intellectual property law did not really exist in Yugoslavia). The whole process was so involved, that the team behind the project published an article in the Review of the National Center for Digitization about their effort.

The results are impressive. Računari magazine never looked so good, and you can tell that the whole project was a labor of love and great attention to detail and historical appreciation. In the countless hours I’ve spent researching the history of the Galaksija, I almost exclusively stared at a screen, looking at PDFs, even after I was fortunate enough to obtain the original first issue of Računari u vašoj kući. That physical copy is quite fragile and I would like to preserve it as much as I can. Having this 40th anniversary reprint lets me read through a physical copy without worrying about damaging a historical artifact.

Here are a couple of examples that show the original and the reprint next to each other. Note how closely the reprint follows the original, with subtle but significant improvements in quality!

So, how can you get one? Luckily, this re-issue is finally ready after years of preparation! You can order a copy from the PC Press webshop – make sure you use the “za inostranstvo” (for orders abroad) issue if you’re ordering outside Serbia. They take all major credit cards and delivery to the US was quite fast.

Galaksija 2024

As you read through the accompanying Računari u vašoj kući 40 years publication, you’ll notice that this isn’t just a re-print project. With the same ethos they embodied in 1983, Voja and Dejan included another DIY Galaksija project to mark the 40th anniversary! Galaksija 2024 is unique among modernized vintage computers in a few ways. First of all, it is not a clone or a recreation — it’s a slightly updated design by the same designer of the original: Voja Antonić. It’s also a pretty unique case in terms of technical updates. While a lot of contemporary projects that bring convenient modern features to old computers use modern microcontrollers and components that contain orders of magnitude more computing power than the computers they’re meant to be used with (and there’s nothing wrong with that!), Galaksija 2024 is virtually unchanged from the original, other than a few small but significant details that I’ll go over.

Board design

The first thing you’ll notice is that Galaksija 2024 is compact — it’s basically condensed to the footprint of the original Galaksija keyboard. This wasn’t achieved through the use of SMD parts or FPGA or some other modern invention. It’s primarily the result of using a two layer PCB and some very efficient connection routing and parts placement performed by Voja Antonić. This isn’t the first time Galaksija has been updated with a second layer. Voja released a gerbers for a 2 layer Galaksija back in 2020, but this was virtually the same design, updated to use Cherry MX style switches and sparing the builder from having to solder 119 jumpers required by the single board Galaksija. With this new iteration, Galaksija becomes a truly compact Z80 computer. Paired with a small monitor, it almost becomes portable! I’m curious if we’ll see any cyberdeck builds of the new Galaksija.

ROM and RAM

The second change is also related to the board design. In 2024, 2732 and 2716 EPROMs are becoming increasingly scarce and expensive, and even when you can obtain them they are quite difficult to program with modern EPROM programmers. Galaksija 2024 is instead built with two 27C256 EPROMs. This helped save space on the board, and also provided additional ROM space. One 27C256 is used for the code from the original ROM A and ROM B + some new functions, while the character generation ROM is stored on the second one (condensing them to a single chip would require additional address decoding logic that would present a significant design change). Additionally, the 3 original 6116 RAM chips are also condensed onto a single CY7C199 32K RAM chip. This also helps save space, and it opens up significant RAM unavailable on the original Galaksija with only 6K of memory. Galaksija 2024 supports a total of 30KB of RAM. These changes however don’t require any significant updates to the hardware design. The only differences are present in software and the memory map, which looks like this on Galaksija 2024:

These changes open up significant RAM and ROM space, but they do come at the cost of losing the expansion port space both in the memory map, and on the board itself. The original Galaksija had an expansion port with direct access to the bus, and a handful of expansion projects were developed for it back in the day. Considering the high cost and limited options of developing traditional expansions for the Galaksija (to my knowledge, none have been developed in recent decades except for those used with the Galaksija Plus clone), the decision to drop support for the expansion port makes sense. And the reason it makes a lot of sense is that Galaksija 2024 introduces a new and much more flexible way to interact with other devices…

EEPROM

I mentioned that Galaksija 2024 uses the same through hole components from the original. This is mostly true, with one exception. One quality of life improvement Voja introduced on this new Galaksija is a small 4Mb M95M04 EEPROM. Since this component is only available in SMD form, it is the one exception in the design that is otherwise fully true to its 1980s form. Using an EEPROM is fully in the spirit of this computer, and unlike flash memory that often gets used as storage replacement on modern retrocomputing projects, this EEPROM runs on the built-in 5V voltage (supported range is 1.8V – 5.5V) and doesn’t require any special decoding logic. But what is the point of adding this EEPROM to the Galaksija? There was one more small change that makes this computer very usable in 2024…

Serial port

Galaksija 2024 implements a TTL serial port! It features a small UART running at 19200 baud and 3.3V, fully implemented in software with the bitbang technique. The serial port effectively functions as the replacement for the cassette interface and the expansion port in one. With changes in ROM, Galaksija programs can now be loaded via serial in a few seconds. It supports the popular GTP format, and programs can either be sent directly through your favorite terminal emulator or via a python based utility that will soon be released. Once a program is loaded onto the Galaksija, it can be saved to the EEPROM. This way you can build a whole library of programs and easily start them from the EEPROM without a need to connect to a real computer to load a program. I’ve been very fortunate to get to test Galaksija 2024 while the ROM and the utility programs were still in development and I have to say that using a Galaksija with the combo of serial + EEPROM is pure joy. You can now load programs in seconds and actually enjoy using the computer, instead of fiddling with .wav files pretending to be cassette tapes (at least that’s how I’ve loaded most of my Galaksija software until now). 99% of historic Galaksija programs in .gtp format work without issues, but there are a few titles like Superšah that implemented copy protection and don’t yet play nice with loading over serial. I’m sure that the community will figure out how to handle a couple of those edge cases as soon as they start playing with this new Galaksija.

Loading programs through the serial port is only the start. The new ROM implements TX/TY and RX/RY functions, so a whole world of hardware integrations will open up soon. Microcontrollers, terminal emulators, only the proverbial Z80 sky is the limit! This is the feature I am excited for the most on Galaksija 2024 and I can’t wait to see what folks will come up with.

Miscellaneous details

In addition to the major changes outlined above, the new Galaksija features a few smaller updates, all within the scope and spirit of the original, which make it better suited for use in the 2020s.

The first is the use of a 6Mhz oscillator instead of 6.114Mhz in the original. This brings the image generated by the Galaksija fully in line with the PAL standard and makes it much easier to use with modern monitors with composite support, or cheap RCA to HDMI converters. Računari u vašoj kući 40 years publication goes into a lot of detail explaining both how and why this works. For anyone interested in the intersection of analog video and digital logic, the article Voja wrote about Galaksija 2024 is a fantastic resource, written with the same wittiness and easy to understand language of his original instructions 40 years ago.

As with the 2 layer Galaksija from 2020, this new Galaksija uses Cherry MX compatible switches (instead of the impossible to find IEVT TY switches from the original). Interestingly, there are some fun anecdotes in the Računari u vašoj kući 40 years publication about how the original Galaksija ended up with the STOP LIST and REPT keys — they were simply the only ones available from the manufacturer at the time. These days things are very different – there are countless keycap options, but very few of them feature some of the antiquated functions featured on the original Galaksija keys.

Getting a Galaksija 2024

Edit: as of October 2024 you can get a gerbers file for Galaksija 2024 at https://racunari.com/galaksija

So how do you build one? The Računari re-issue features detailed instructions for building Galaksija 2024, but since the hardware design was still going through final revisions, it doesn’t feature the actual PCB design. This is not a problem, since these days we manufacture PCBs from gerbers files, not photocopies from printed magazines (even though maybe we should?). PC Press just announced that they are taking orders for Galaksija 2024 kits through their webshop. Check racunari.com for more info.

Unfortunately, due to customs regulations in Serbia, they will only ship domestically (within Serbia) for now. There is likely a whole another article to be written about how the 40th anniversary Galaksija struggles with the same problems the original one was built to overcome, but I will leave that for another time. For those of you in the EU, the US and elsewhere, I recommend you watch the racunari.com page for announcements on when they might start shipping globally. And make sure to check out the racunari.com/galaksija page which features the schematic, the parts list and other downloads. It will feature the new ROMs soon, and Voja will be releasing the gerbers as well Both the new ROMs and the gerbers files have been released, so everyone can build one by going to their favorite Chinese PCB manufacturer. While you wait, make sure you check out the video on PC Press YouTube channel where Dejan Ristanović unboxes a Galaksija 2024!

Conclusion

With all the recent round anniversaries in computing, I’m so happy to see this project celebrating the Računari magazine and the Galaksija computer, lead by their original creators. Especially given the long period of time when they weren’t given the credit they deserve! The continuation of the public interest in the Galaksija 40 years later shows how its story is still relevant, and this project makes participating in that story more accessible. It’s been a great pleasure and privilege to get to test out the development version of the new Galaksija and I can’t wait to see what people do with it!

The post Galaksija and Računari u vašoj kući: 40th Anniversary Celebration appeared first on Vlado Vince.

It’s the middle of MARCHintosh! Thanks to the unexpected GlobalTalk craze, I spent the first two weeks of March having a great time participating in the world’s first global AppleTalk network. The first few days were wild — I’ve already documented my setup to interconnect a mini vMac emulator to my SE/30 running the Apple Internet Router to connect to the rest of GlobalTalk, but as of March 16 this post is already out of date. Everything I shared there is still valid, but I’ve made some pretty big changes to how my local AppleTalk network is set up, and how it connects to others in GlobalTalk, so a new post is in order. The goal of this post is to document my setup, and hopefully help others get OnLine on GlobalTalk.

If you’re not familiar with GlobalTalk, I recommend checking out the “official” MARCHintosh GlobalTalk page as well as Paul Rickards’ Pushing AppleTalk Across the Internet post, which helped bring many folks on board in early March (including me!). GlobalTalk got started on Mastodon in February when a couple of folks (Paul and dan) started brainstorming about connecting their LocalTalk networks across the Internet. To look up that “early” GlobalTalk history, search for the #LiveLaughLocalTalk hashtag on Mastodon:

Their first attempt was to use AppleTalk VPN, but by late February they pivoted to Apple Internet Router. This was the secret ingredient that allowed rapid expansion of the network. With a list of participants’ public IPs, AIR pretty much automagically configures tunnels to remote AppleTalk zones. The most complex part of the setup is configuring your home router’s UDP port 387 outside to UDP port 387 inside to the IP address of your Mac running AIR. By now there are tons of different ways to run AIR — from 68K and PowerPC Macs with Ethernet cards and even BlueSCSI, to emulated solutions using QEMU or Basilisk II. Client machines are even more varied – there are two decades worth of Macs that support AppleTalk, from the earliest Macs up until OSX Leopard. Folks have connected Apple Lisas, Apple IIGSs, all sorts of printers (including multiple *cat printers*), Apple Newtons (inlcuding *printing from them*), countless RPis and Linux boxes running Netatalk 2.2 and even Windows NT!

In the last few weeks there have been countless creative and interesting uses of this unique anachronistic network, and GlobalTalk deserves that a proper oral history be written down. However, since MARCHintosh is still ongoing, and folks are still coming up with new ideas, I’m going to leave that for another time. My goal with this post is to share a bit about how I built the Apple Internet Router portion of my GlobalTalk zone which is called JUPAK. If you’re not on GlobalTalk (yet?), but you’re curious about what’s going on, search for #GlobalTalk on Mastodon.

Apple Internet Router (3 phases)

Phase 1: SE/30 with BlueSCSI v2

The Apple Internet Router is (currently) the core of any zone connected to GlobalTalk. I initially set up my Apple Internet Router on my recently rebuilt Macintosh SE/30. This was the early days — at this point everyone was running their AIRs with Ethernet cards, and I didn’t have one (yet!). We weren’t yet sure if solutions like Daynaport emulation on BlueSCSI v2 would work, so I first considered connecting via PPP (which would likely be unbearably slow)

Luckily, very quickly after this I was able to get AIR connected to the rest of GlobalTalk using the aforementioned BlueSCSI v2! At this point I was using a copy of an already configured image from a Quadra 800 running AIR on System 7.1 (prepared by Paul Rickards). The setup was very straightforward — I had to adjust AIR to use the emulated Daynaport as its Ethertalk interface (it would show up as just Ethernet in my AIR) as well as the interface for the AppleTalk connection in the Network control panel, where it shows up as EtherTalk Alternative. This was enough to get me on the network and I was able to start having fun! At this point, printing to other folks’ ImageWriters was all the rage, so I immediately got on board too:

This setup worked, but I very quickly realized that with virtually no antenna and a 2.4Ghz connection, my BlueSCSI v2 was struggling and dropping packets left and right. As a temporary measure, I set up an access point literally next to the SE/30. This improved the connection considerably:

But this was still just a temporary setup, because I had an Ethernet card on the way…

Phase 2: SE/30 with a SEthernet/30 card

I was incredibly lucky that Richard Halkyard’s SEthernet project of building a brand new Ethernet card for the Macintosh SE and the SE/30 wrapped up pretty much in time for MARCHintosh, and I was even more fortunate to be one of the folks to test their new card!

SEthernet is a brand new design that solves an all too common classic Macintosh cost conundrum – due to their current high demand, original Mac expansion cards sell for hundreds of dollars, while clones of those cards often use parts that are difficult to obtain. Richard solved this issue by designing a brand new Ethernet card around a readily available controller!

The card is built around the Microchip ENC624J600 all-in-one embedded ethernet controller. It’s cheap and available, and it plays nice with the 68k architecture. There are two versions of the card – the SE and the SE/30 version. Due to limited expansion infrastructure on the SE, it lacks some of the features of the SE/30 version.

While this is still technically a test board, Richard is gearing up for production and I recommend that you follow them, as well as the SEthernet GitHub page and the 68kmla project thread.

Once I popped SEthernet/30 into my SE/30 and installed the driver Richard wrote, I just had to adjust my AppleTalk interface to use the appropriate interface (EtherTalk (Slot 1)) and rebuild the IP list in AIR, corresponding to Ethernet (Slot 1)) interface.

As expected, a proper Ethernet card significantly improved my overall performance on the network! This also coincided with me getting zigzagjoe’s Booster 50-SE30F accelerator which increased the clock speed on my SE/30 to a whooping 47Mhz! The SE/30 was now a formidable router.

Phase 3: emulated router in QEMU

Once I had the SE/30 properly accelerated and equipped with an Ethernet card, my connection to GlobalTalk was rock solid. However, since most of my collection of eclectic AppleTalk machines primarily uses LocalTalk, I wanted to use my spiffy SE/30 as a client machine for chatting with folks, playing games and exploring the weird things they’ve made available on their networks. Since I didn’t have another good candidate that would run AIR (for a second I considered running it headless on another recently rebuilt SE/30 board but luckily decided against that), I decided to emulate the routing function.

While I managed to successfully run AIR on Basilisk II with UDP forwarding, such configuration can’t support routing for other machines on the network. Right around this time, europlus preparred a pretty-much ready to go package that included a pre-configured System 7.1 image and commands to run AIR inside QEMU with a properly bridged network adapter. I initially wanted to run this from a Linux host (and I still might at some point!), but QEMU network interface bridging seems to be a lot more straightforward in current macOS than it is in Linux, so I decided to dedicate a 2019 Macbook Pro to AIR duty for now.

On macOS Sonoma the setup is incredibly straightforward. I got the image and startup scripts from europlus, and the only thing I had to adjust was to configure the startup commands to use my USB ethernet adapter as the network interface. And I had to update the NAT on my router point UDP port 387 to the now emulated machine. (There was also a brief nVIR virus scare where we found out that some of our system images got infected by a 30 year old virus, but you just had to be there to fully appreciate that one.)

This is all somewhat ridiculous — I’m now using a significantly newer and more expensive machine to perform this simple task. However, it does come with some quality of life benefits — I feel comfortable leaving it on all day without fear of it catching fire or breaking down (if you know the story of my SE/30 you will understand my paranoia). This means that others are able to access my shares and resources even when I’m not home. Additionally, I have easy remote access via VNC, though QEMU does have some struggles with mouse movements that take some time getting used to.

One modern struggle I have with this setup is macOS Sonoma’s insistence on putting the computer to some sort of sleep as soon as the lid is closed. I tried to overcome this by using Amphetamine but no matter what I tried the laptop still went to sleep at some point after the lid was closed. So for now I’m just keeping it open.

Next steps

I’m pretty happy with my current GlobalTalk setup, at least in terms of AIR. I don’t know how long this experimental network will last — I expect that many folks will start to disconnect once MARCHintosh is over, so I’m trying to have as much fun as I can with it right now. In terms of documentation, this is the first of a few articles I plan to publish. In the upcoming posts I will cover my AppleShare setup with a netatatalk 2.2 docker instance, as well as my LocalTalk bridge setup (including bringing my 512K online!) with the help of tashtalk and tashrouter (effectively an update to my initial setup I documented a few weeks ago). I also plan to document some of my mini projects, like the GlobalTalk hosted README file that’s accessible at https://jupak.vladovince.com, as well as my Spectre Multiplayer status page which shows currently connected players playing Spectre over GlobalTalk.

The post My GlobalTalk Setup: Apple Internet Router configuration appeared first on Vlado Vince.

It’s MARCHintosh! The month when the vintage computing community celebrates the classic Macintosh experience. This MARCHintosh started with a bang – seemingly overnight, the #GlobalTalk network took over Mastodon and soon many of us were networking each other’s disparate home AppleTalk networks into a #GlobalTalk network across continents. We’re still in the middle of the #GlobalTalk craze and I expect to write more about my experience further down the line, but I can already say that this is by far the most fun I’ve had going “On Line” since I was a kid in the late 90s. If you’re curious about #GlobalTalk, look up the hashtag on Mastodon and get started with Paul Rickards’ blog post which documents the essential info to get started. Also check out Andy Diller’s blog for some quick start archives that contain all the software necessary to connect to #GlobalTalk.

Bridging LocalTalk and LToUDP with the help of TashTalk

One of the challenges working with classic Apple networking standards is bridging connections across different networking interfaces, standards and layers. The challenge becomes even more interesting when we want to facilitate communications between real computers and classic Macs running in emulators. Luckily, nowadays there are a number of software and hardware projects that can help. Once I got connected to the #GlobalTalk network using my SE/30 as the Apple Internet Router, I wanted to connect to the network from my Linux PC by using an emulator. I chose Mini vMac due to its support of LocalTalk over UDP (LToUDP) in version 37 beta. However, I first needed to turn the normal LocalTalk connection, accessed over the ModemPort, into LToUDP — this is where the TashTalk comes in.

TashTalk 2 Hat is a very simple Raspberry Pi hat that handles turning LocalTalk serial signals into multicast UDP packets that will be picked up by LToUDP enabled clients on the local network TashTalk connects to. TashTalk was developed by tashtari over at 68kmla forums.

Like many other retrocomputing projects, TashTalk has a hardware and a software portion. The hardware consists of a pair of chips (RS422 transceiver and a PIC12F1840), a capacitor, a 40-pin header and a mini DIN connector. I bought a kit from tashtari and it came with a pre-programmed PIC12F1840. All I had to do was solder the components to the board, which only took a few minutes.

TashTalk 2 Hat mounts on top of a Raspberry Pi. I tested it with both a Raspberry Pi 2 and a Raspberry Pi Zero 2. Reportedly TashTalk has issues with WiFi, so I installed mine on a RPi Zero with a USB ethernet dongle.

Once you have your hardware assembled, you need to set up the software portion. You will need to place an operating system image on the Raspberry Pi. I used a regular Raspberry Pi OS (formerly Raspbian) image. I connected the Raspberry Pi to my local network over Ethernet, and to my SE/30 with a LocalTalk dongle (note that per tashtari you have to use a LocalTalk or PhoneNet dongle, a serial cable will not work due to the type of transceiver used).

I then proceeded to get the TashTalk software from github by cloning its repo to the Rpi. I then navigated to the tashtalkd folder and ran the program with ./tashtalkd --device /dev/ttyAMA0 -vv &

-vv is just here for added verbosity. Since I was already physically connected to the SE/30 I immediately started seeing LocalTalk traffic:

This was fantastic — the TashTalk just worked! The next step, getting Mini vMac to work, was a bit more difficult. Note that you will need to add the tashtalkd command to some kind of startup script unless you want to run it manually each time.

Connecting Mini vMac emulator to SE/30 via TashTalk and LToUDP

Instructions to use a Mini vMac with Localtalk over UDP. These instructions are for generating a beta version of a standard Mini vMac which emulates a Macintosh Plus.

Generate a localtalk enabled Mini vMac executable

• download Mini vMac 37 (Beta) from https://www.gryphel.com/c/minivmac/beta.html
  • it’s important to get the beta version because standard version 36 only supports localtalk on OSX
• follow instructions to build vMac: https://www.gryphel.com/c/minivmac/build.html
  • first compile the configuration tool gcc setup/tool.c -o setup_t
  • generate your setup script with ./setup_t -br 37 -t *YOUR ARCHITECTURE* -lt > setup.sh
    • for example, a Linux 64 bit is generated with ./setup_t -br 37 -t lx64 -lt > setup.sh
  • make your setup script executable (chmod +x setup.sh)
  • execute the setup script ./setup.sh
  • compile the executable with make
  • this will make an executable called minivmac
• place the minivmac executable in a folder with a copy of a Macintosh Plus ROM. You should source this yourself and name it vMac.ROM
• you will also need a working system with AppleTalk support. You can set this up yourself or use a pre-set image like some that you can find here: https://macintoshgarden.org/apps/mini-vmac-ready-go
  • you may also find the ROM in some of those files
  • I used version Mini_vMac_os_7.5.3_ready_to_go.zip
  • make sure you use the Mini vMac executable you generated above with localtalk enabled, not the one attached to the ready-to-go image!

You can also skip manually building your executable. Gryphel.com features a Variations Service where you can build your own vMac based on requirements. Note that by default this service will add a “demo” badge on your build, but you can remove it by the use of a code that is now published on the intro page. If you’re using a 64 bit Linux host, these are the settings you should use for your image:

Here’s a list of all the codes for different host platforms:

{Beta} -t mcar { Macintosh OS X - ARM (Apple Silicon) }
-t mc64 { Macintosh OS X - x86-64 }
-t imch { Macintosh OS X - x86-32 }
-t mach { Macintosh OS X - PowerPC }
-t wx64 { Microsoft Windows - x86-64 }
-t wx86 { Microsoft Windows - x86-32 }
-t lx64 { Linux - x86-64 }
-t lx86 { Linux - x86-32 }
-t lppc { Linux - PowerPC }
-t larm { Linux - ARM }
-t lspr { Linux - SPARC }
-t fb64 { FreeBSD on x86-64 }
-t fbsd { FreeBSD on x86-32 }
-t nb64 { NetBSD on x86-64 }
-t nbsd { NetBSD on x86-32 }
-t oi64 { OpenIndiana on x86-64 }
-t oind { OpenIndiana on x86-32 }
-t wcar { Pocket PC - ARM }

Once you have your Mini vMac executable, start it with your image. If your install is ready for AppleTalk, navigating to Chooser will immediately show you all the #GlobalTalk AppleTalk zones your Apple Internet Router computer is connected to:

And that’s it! You can now participate in any of the fun activities folks are doing on the #GlobalTalk network, such as printing to an ImageWriter on another continent, chatting inside a HyperCard application or playing Bolo with friends in different countries.

Note: this is not the only way to join #GlobalTalk from an emulated Mac. I will also publish the instructions on how to use Basilisk II to run your own Apple Internet Router from an emulated machine (though for now I don’t think it’s possible to connect it to other physical machines)

The post Using TashTalk with Mini vMac to participate in #GlobalTalk from your emulated Mac appeared first on Vlado Vince.

In early September I published a post on Mastodon, with what was meant to be a humorous comment. “Oh no”, I said, implying that there is something wrong with getting a highly coveted Macintosh SE/30. The joke was accompanied by a photo of a small beige Mac on a Long Island Rail Road platform. If only I knew how many times I would repeat that “oh no” over the following few months.

At the time, the “oh no” mostly referred to my concerns about bringing yet another largish vintage computer into a small Brooklyn apartment. We had recently moved, and I hadn’t even unpacked all my existing old crap. Plus, this would be my third compact Macintosh, joining a regular SE and a 512K I restored recently (with quite a lot of trouble). In fact, I was so concerned about space and so sure that this restoration was going to be quick and painless, that I immediately posted about looking to find a new home for the SE. Oh how quickly I would learn that storage would be the least of my issues with the SE/30.

The two boards

Considering the context, it’s not surprising that this was an impulse purchase. It was the Labor Day weekend, and around noon on a Monday I came across a post in one of the vintage Macintosh Facebook groups where someone posted a photo of a wall of compact Macs, looking for buyers to take them all. I immediately noticed a single recognizable SE/30 among a few SEs and Pluses. And then I noticed the location — just a train ride away from me. While the seller offered to ship items as well, I knew that heading over asap was the only way to ensure I would get it, dangers of shipping compact Macs notwithstanding. I quickly got in touch and confirmed with the seller that I would be there in a few hours. A relatively quick train ride later, I showed up and got to inspect the Macintosh I would be taking. It was immediately obvious that there had been battery leakage in the case – the seller allowed me to open it up and we confirmed that the board was badly damaged. While I hesitated, they offered to include another SE/30 board, in great physical shape and recapped as well, for the price of $100 total. I had already spent a bit of time and money to get there, so I gladly agreed and took the SE/30, and the extra board, back to Brooklyn.

Board #1 — the battery bombed disaster

The board that was inside the case was disgusting. The photos included here were taken when I got home, but by that point some of the most visible dirt and dust was gone (see above how it looked like originally). It was clear to me, even with my limited experience with battery damage, that there should be no attempts to turn this board on without serious cleanup efforts. The battery holder fell off the board, some bottom caps were gone too, and the battery (and capacitors too) have started to do damage to a lot of components.

Board #2 — the misleadingly good looking catastrophe

In comparison, the second board the seller gave me looked great! It was clean, with no visible damage. Even the capacitors were replaced, and whomever did it did a very clean job. I though I had hit the jackpot — as long as the analog board in the case was fine, this would be smooth sailing.

The smoke test

Before venturing to test board #2 (I wouldn’t even consider plugging board #1 before a lot of cleanup), I went ahead and checked the analog board, PSU and the monitor. Plugging my known-good SE board, it booted right up. Good sign!

This optimism continued for a while longer, as I connected board #2 and got the floppy disk icon! Things were looking up, until I tried to boot from a BlueSCSI which didn’t even get detected. While that was a bit discouraging, I pulled out my FloppyEMU and tried to boot into a system. The system would start loading, but then fail with a flashing screen:

This happened with any system I tried. So there may be some work left to do, I thought. I couldn’t predict just how much work.

The early days of repair

I’ve been repairing electronics for as long as I can remember, but I really got serious about vintage stuff about 3-4 years ago (remember the still ongoing pandemic?). During this time, I’ve built a couple of 8 bit computers and restored quite a few more. I even designed a few hardware projects from scratch. And this was my third compact Mac — I expected a few days or two weeks of work max to get this resolved. I started with the easy stuff. Everyone says you should recap your Macs, so even though this board was clearly recapped, I went for it anyway. This did absolutely nothing, the SCSI still wouldn’t work and booting into any system failed immediately.

Checking connections on the SCSI chip wasn’t helpful — the 53C80 chip had good connections to everything. Since the chip was still available on Mouser, I got a new one and replaced my old chip. This was the beginning of my long and arduous road of teaching myself how to properly solder and desolder surface mount components. Yes, everyone says get yourself something worthless like a DVD player and practice, but I was:

a.) Impatient

b.) Convinced that I already knew how to do it based on limited previous experience

Looking back, I absolutely had no idea what I was doing (as will become apparent later on). I was aware of the danger of damaging the board when removing components (I did have some experience), so my first attempt at removing the SCSI chip included the use of chipquik, a low melt solder that stays liquid long enough to remove a component. I combined this technique with the use of a heat gun (yikes) and I managed to lift a pad. I did make a small bodge connection to the trace it connected to, but this was a good indicator of issues to come.

The good news was — SCSI actually came to life! The computer would attempt to boot from it, but sadly it would again fail to load a system, crashing with a variation of flashing welcome screens. At this point I came across a video that gave me hope for a moment.

In this video, Joe comes across an issue that looks the same way as what I was experiencing. Flashing screen that wouldn’t complete system loading. He then manages to boot into System 6.0.4 and concludes that the issue is related to the FPU. He bodges the 5V rail to the CS line on the FPU, effectively disabling it. Afterwards he’s able to boot into other systems as well. This got me excited, and lo and behold I managed to boot into 6.0.4 as well!

The excitement was short lived. The system immediately crashed, and subsequent attempts to boot it failed as well. I tried the CS trick to disable the FPU without success. This seemed like a red herring. But it did convince me that there was some connection to the FPU. I placed an order for a replacement. I also ordered a Rominator II on the off chance that the issue I had was related to the ROM. While I was at it, I also ordered replacement F258 and F253 muxes. By this point I learned about their role in the memory bus and I convinced myself that the issue must lay somewhere among them. While waiting for all these parts to arrive, I tried every possible memory SIMM combination, with no success. I used confirmed good SIMMs, in combinations of 4, failing each time in the same way. The problem was somewhere else.

The first to arrive was the Rominator. It immediately failed in the same way. Then I tried the F258 and the F253 muxes. Same issue. I then replaced the FPU. Same issue.

Wait. No. Not just the same issue.

Collapsing image

A few minutes after soldering on the new FPU, the real issues began. I started to experience what I’ve been calling the “collapsing image” problem. At this point, I began to panic. Instead of taking it slow and thinking this through, I started to make assumptions and switching out components from the other board (somewhat cleaned up by now). Is it the PALs? I started to program new GALs. Is it the CPU? I even replaced the 68030, a very difficult task considering its 132 pins.

At this point I’m making mistake after mistake by reacting without a plan or process. Looking at the schematic, I very early on hone onto the UH7 chip which takes the 32Mhz clock. I observe that clock line collapsing, and see the effects on the screen. Early on the computer chimes, first a good chime, then the bad. Soon, the chime completely stops. Sometimes the jail bar (one version of simasimac) shows, sometimes for a few minutes until it collapses without fail. The bus is dead.

Cleaning up board #1

At this point I’m going to do a short detour. Before completely blowing up my repair with the collapsing image problem, I worked on the battery bombed board (board #1), cleaning it up and eventually trying to re-connect broken lines. While this effort was ultimately completely unsuccessful, it was useful as it cleaned up many components that I would later on successfully reuse. I started by submerging the board in vinegar and leaving it overnight. I then scrubbed it with a toothbrush. Finally, I bathed it in IPA and washed with distilled water.

The results were not horrible:

In retrospect, I was able to save most components. The exception were the SCSI chip, the FPU, the ROM SIMM and some of the video logic PALs that I damaged while desoldering them.

At this point I also tried to recover this board by reconnecting broken traces and replacing disintegrated resistors and capacitors on the bottom. This never progressed past the simasimac (jail bar) pattern.

Once I encountered the collapsing image issue on board #2, I first used board #1 as the donor for UH7 chip. I flipped the chips between the two boards, and suddenly, board #2 which wouldn’t boot but at least had the stable jail bar pattern now started exhibiting the same collapsing image symptom!

Moratorium and decision to rebuild

After the initial collapsing image problems started and I had no success resolving them, I had to take a break. I gave myself a 2 week moratorium during which I wouldn’t touch either board. Instead, I started looking into alternatives. I was sure that between the two boards I had enough good parts for at least one working SE/30. Additionally, with all the removal of components from board #2 I was getting worried that I am just damaging it further. By this point (early October) I had lifted a few pads and without proper tools (I was still using a heat gun!) I was likely damaging traces inside the PCB as well (the SE/30 uses a 6 layer board!). I arranged to get an SE/30 reloaded board, designed by Bolle from 68kmla.org forums. I arranged a single board from a fellow enthusiast, and they threw in some hard to find components like the RAM and ROM sockets and the PDS slot. I placed a large Mouser order for as many other replacement components to minimize the amount of items I would need to remove from old boards.

Parts harvesting

I decided to get my parts from board #1 (battery bombed one). I still had some hope for board #2, and I felt that destroying board #1 in the process of component removal would be acceptable. The difficulty of removing the parts varied. I finally went ahead and got a hot air station (something I should have done before ever touching an SE/30) and with enough heat any soldered component will eventually come off. The biggest problems are sockets with plastic parts. Luckily for this first new board I had no need for plastic components. The most difficult piece was the metal “frame” which clips into the chassis. By the end the board was full of scratches.

Board #3 – Reloaded attempt 1

I got my reloaded board in mid October. It came with the bottom pre-populated, sans the ferrite beads which I sourced from the donor board. I was excited and determined to get it right. I had all the parts ready. I had fresh soldering tips and a jar of flux (remember this foreshadowing for later). I set up music and embarked on populating the board. I took ample photos documenting the process, and I completed the build in some 8 hours. I flipped the switch.

Not only did the board not boot, I had an actual visibly detectable short somewhere. It didn’t even pass the smoke test.

Months later someone pointed out that in the image above I actually have an incorrectly soldered IC. Look at the ADB chip. It should be rotated by 90 degrees left. I soldered it like that because I was using board #2 to look at chip orientation, and that particular ADB chip on that board has the part number rotated.

I immediately lost steam. I spent days tracing the short, and frankly I don’t think I ever actually found it. But I did get the board to a state where it did the exact same thing as the other two boards — the image was collapsing after a few minutes.

Rebuilding the clock and video sync circuit

At this point I gave up the dream that this was going to be quickly resolved. I tried the analytical approach to just get a stable image. I had no intention of getting this to boot. If I could get this board to hold an image without collapsing, then I would proceed.

So how does the SE/30 generate video sync? Since we’re dealing with analog video, we need a horizontal and vertical sync. We start from the Y2 32Mhz oscillator. This is the main crystal that is also used for the 68030s 16Mhz clock. Y2 provides 32Mhz to the UH7 PAL. This PAL then passes that signal onto both the UI8 GLUE chip and the UI6 PAL as C32M. The GLUE chip splits it in half to 16Mhz and passes it to UI16 as C16G, which then passes it onwards as C16M. Then additional video logic PALs, in conjunction with LS393 binary counters, generate VSYNC and HSYNC signals (UG7 for HSYNC and UG6 for VSYNC).

This, roughly, is all the logic needed to generate video sync signal (there’s lots more passive components necessary). Without adding anything else, you could get an image like this on the CRT:

If you add the video RAM, 74LS166 shift register and the remaining PAL chips, you’ll get the jail bar pattern (with the particular video RAM I am using. Other types may give other simasimac patterns):

What I observed on board #3, as well as the other two boards, was that the video sync signal starts to break after a while, resulting in the collapsing image symptom. It will start with VSYNC and eventually collapse HSYNC as well. At this point I suspected a bad component — perhaps it was the oscillator?

I replaced the oscillator.

Perhaps one of the PAL chips?

I replaced the PAL chips.

Perhaps the GLUE chip?

I couldn’t replace the GLUE chip since it’s a custom chip that hasn’t been reverse engineered yet. But going back to the video sync generation logic, we see that the only use of the GLUE chip for generating the signal is to divide 32Mhz in two. I could do that without it.

So I built myself a frequency divider, and programmed a 20 pin F16V8B GAL instead of the UH7.

Same. Damn. Thing.

I was hitting the same wall again and again and again.

Power theories

For a while I even suspected the power supply. I had the same issue in my SE chassis, and my SE board would work fine in the SE/30 chassis. I even wired up an ATX power supply. I recapped the analog board.

Someone suggested my electrical wiring at home might be bad (!). I didn’t really think that could be the case since the building is less than 5 years old, but sure. I ran the power supply off of a UPS.

Same. Damn. Thing.

Giving up

This was going nowhere. I somehow replicated the same issue on a brand new board, and considering the shorts, it was unlikely this thing was going to be recoverable. By this point I spent literally countless hours and quite a bit of money on this project, and I was actually in a much worse place than when I brought the damn thing home. I was disappointed, and the worst thing was, I felt like I didn’t even learn anything.

At this point while folks were still encouraging me to continue, I really felt like giving up. Helpful people shared advice, but we were past the point where this could be troubleshot remotely.

One thing that kept coming back was the idea that there was a short somewhere. But I couldn’t detect it. The fact that there is relatively small resistance between ground and 5V on a good SE/30 board made this difficult to prove.

I left this for a few weeks. Eventually, I made a decision I would give it one, final try.

Regrouping

I ordered a set of brand new reloaded boards.

JLCPCB has a minimum of 5. The cost was more than I paid for the SE/30 and the two boards in the first place ($150 with shipping). At this point this was either going to work, or not. I would use these boards to experiment, to finally crack how this thing works. It wasn’t about getting a functioning SE/30 anymore — it was about getting a small set of electrical components to behave well enough not to collapse.

Boards #4 and #5

I was patient, and precise. Slow and methodical. No unnecessary components. Minimum viable circuit. Despite some early stability, two times I tried again, and two times I failed. Boards #4 and #5 were doing the same thing as all the other ones.

I stepped away again. I thought about everything that’s gone into this. I couldn’t shake the feeling that I was somehow infecting these boards with this issue. There was barely anything common between them. Yes, I reused the GLUE chip, but I also got the same stuff with my own clock divider. Something else was causing this. Something else was the pattern.

Oh.

This was such a relief. I didn’t yet know what the actual problem was, but it made sense. A total of five logic boards, same symptom. I replaced everything else, other than the person doing the work. It should have been obvious earlier on, but the one factor I did not control was the actual work being done to solder different components onto these boards. It was me – I was the pattern.

I took another break for the month of December. I traveled and decided I would continue in the new year. I had 3 boards left, and if I couldn’t get it right with those, I would pass this chunk of SE/30 parts to someone who can. I thought about what I could be doing wrong, and the only thing that came to mind was the soldering of tight SMD chips. This all started with me removing the GLUE chip on board #2, and even for the minimum viable video sync circuit I had to at least install UH7. So let’s talk about UH7 for a second…

UH7 passes on 32Mhz clock signal from the oscillator. It’s also the only PAL logic chip that’s in PLCC-20 instead of DIP-20 form. It’s located in between the oscillator, the GLUE chip and other PAL logic. It’s hard to access. And it’s pads can barely be accessed with a soldering iron. It’s tough to solder.

Epiphanies

Even when I replaced the chip, as in the image above where I’m using an ATF16V8C, I would have to solder it on. What could I do to ensure my soldering wasn’t making tiny shorts? A friend mentioned that I should try to use sockets — they might be easier to solder, and I would be able to test different components more easily. I initially decided against SMD sockets because of concerns that they would cause contact problems (this may be foreshadowing something later). But I had much bigger issues. In early January I got some PLCC sockets, and decided to test it on board #2, the original sort-of-good board, by placing the socket on UH7.

Also, I got new flux (remember this for later).

Guess what? The image did not collapse. I ran it for 90 minutes, with the sad jail pattern, and the image did not collapse.

I’m the problem, it’s me.

This board obviously still had problems preventing it from booting. After all the work I’ve done, I suspected there may have been broken connections inside the board. I decided to go for one more rebuild, this time with sockets, and with a full arsenal of parts — I would use board #2 as donor as well. It was all or nothing.

Board #6 – The board that boots

I was patient. I was precise. I socketed everything. Yet, I would lie if I said this board booted immediately. However, it never collapsed, and it was showing signs that it’s trying to do something. At this point I tried to be as analytical as possible in my troubleshooting, precisely understanding what the board does in different conditions. Techknight from 68kmla.org was very helpful with his SE/30 videos, especially the one where he demonstrated how the 68030 walks the memory bus when no ROM is installed:

On January 20th I finally got a successful boot.

It was a matter of identifying functioning parts and (as I would later learn) getting them to fit in sockets just right to get them to work.

I was ecstatic! Approximately 120 days after I brought this SE/30 home it was finally fully working. In the last stretch I even got a replacement 68030 (something that later on proved to be unnecessary), but it didn’t matter. This computer Worked.

I proceeded to experiment and play with it for a few weeks afterwards. It was, after all, well earned.

But, it was not the end…

The revenge of the sockets

A few weeks later, I decided to build board #7. At this point I knew I could do it, and I had enough parts for another one. I didn’t hurry too much, but I did switch chips between the two. While still working on board #7, I started to notice strange issue with board #6 (uh-oh). It suddenly started randomly crashing and reporting CPU and FPU issues in Snooper. Checking the machine info showed a strange 2058 Mhz speed which was obviously incorrect.

Soon the FPU stopped being reported at all, and any application needing to use the FPU would fail with Error Type 1. I tried booting without an FPU (since I have a Rominator which can boot without it) and I would get Error Type 90 (i.e. “an FPU instruction was executed but the machine doesn’t have an FPU”). I also had SCSI trouble, and floppy trouble. What was happening? Was I infecting my machines, again?

I had 3 FPUs, and none of them worked. I got another one, and the same thing happened. It must have been the sockets. Like UH7, FPU is tricky to solder because it’s surrounded by plastic components and the pads are very tight. I hesitated to solder the FPU directly without knowing whether it was good. I decided to go with a new socket, but I left the plastic bottom in place (previously I had cut it out). It made it somewhat difficult to solder in the corners, but luckily the FPU showed up correctly in the system immediately after installation.

I also went ahead and soldered the GLUE and SCSI chips directly, resolving their respective issues. After all the help from the sockets, they came back to bite me. Speaking of getting bitten…

Board #7 – The one that confirmed the problems

In February I decided to build board #7, ie. second new board that would work. The build was uneventful, until I ran out of flux. I wanted to keep going, so I dipped into my old jar of flux.

Guess what happened?

Yes, the sockets probably helped, but I can safely say now, my issue with the collapsing image was initially, and subsequently primarily caused by this stuff:

I now know that rosin flux can be conductive at certain temperatures. It makes sense that it would cause shorts that would take a few minutes to show up, as it heats up. In retrospect, I believe that this thing probably caused similarly bizarre issues with a BlueSCSI I assembled a few years ago. Yes, I should have been better about cleaning it up, but this stuff ended up in the trash and I’m never using it again.

After a thorough clean in IPA, washing, another round of IPA, washing and cleaning, and drying of course, board #7 would not collapse, proving once and for all that shorts were the cause of the issue, and my rosin flux was the likely culprit all along.

Getting board #7 fully working took some more work. Sockets ended up being an issue again — I ended up soldering SCSI, GLUE and UK11 (one of the VIA chips). I replaced the FPU socket for one with a plastic bottom because the FPU again wouldn’t get recognized. I even had to replace the socket for the video ROM because the cheap socket I had in there wouldn’t make good contacts with the ROM! This one was weird to troubleshoot and eventually I figured it out by seeing the computer boot from SCSI by its blinking lights when the video ROM was be taken out. It must have been messing with the bus. All in all, board #7 took about 3 weeks from start to fully functional.

Chassis rust removal work

Sometime in the fall I spent a few days working on the badly rusted chassis that was damaged by the battery explosion. I started by giving it a vinegar bath.

After a while, the vinegar cleaned it up really well, but the metal still stained with very thin layer of rust:

This was very annoying. I went through a couple of rounds of vinegar baths until I realized that there was no way I would get all the rust off that way. Moreover, the stains would appear on the parts of the metal that had no rust in the first place! I decided to remove the rust with a dremel and steel wool.

Finally, I applied a few coats of Rustoleum spray. The color is close to the original, but note that this paint is not conductive so I made sure the points of contact with the logic board and screws were scratched enough to make contact.

Before and after:

Cost

I hesitated whether to calculate the cost of this project. This work is clearly irrational outside the very particular context of a weird hobby. While I did not want to buy an overpriced SE/30 on principle, I spent A LOT more money on getting one complete SE/30 and an extra board than I would have if I just bought an overpriced SE/30 in the first place. But, I think it is interesting and useful to share this information transparently like this. Note that this does not include any of the new tools I bought (like the hot air station) nor additional hardware like new RAM, Rominator II and BlueSCSI V2 — I only counted the items I needed to actually get these boards working.

I wouldn’t want to speculate the cost in my time as it would be more upsetting than the collapsing image issue itself. I did learn something in the end, so let’s leave it at that.

Final notes

This project got so blown out of proportion in time, effort and money that it is really difficult to assess whether it was worth it. It challenged me by an order of magnitude more than any electronics project I’ve tackled to date, and I can say that I learned a lot. I could probably rebuild an SE/30 in less than 2 weeks now. I probably won’t, though — this is a hobby for me, and I’ve successfully figured this one out — it’s time for something new. We’ll see what comes next.

I wanted to share some resources that I found indispensable working on this project:

• mishimasensei’s SE/30 schematic recreation
• Original SE/30 schematic
• Bolle’s logic PAL dumps and the whole SE/30 logic board recreation project
• THEtechknight’s videos and posts on 68kmla.org
• Branchus Creations videos
• JDW’s recommendation to use an ATX extension cable to work with your logic board outside the case
• 68kmla.org forums
• TinkerDifferent forums
• Encouragement, tips and comments from many folks!

Repairing old electronics is fascinating. For a discipline that is based on digital logic and the concept of binary states, it’s fascinating how much of working with this technology is almost esoteric. I really appreciate the importance of experience and craft after completing this project. It really was a matter of skill and experience that made a difference in finally bringing this one to a close. Thanks for reading!

The post I get knocked down, but I get up again: my Sisyphean Macintosh SE/30 struggles appeared first on Vlado Vince.

Note: this post is based on a Mastodon thread.

Last summer I got a Macintosh 512K at the VCF Swap Meet in Wall, NJ. I’ve been wanting to get a Fat Mac for a while, ever since I read Andy Hertzfeld’s book Revolution in the Valley: The Insanely Great Story of How the Mac Was Made. Well, technically I wanted a Mac 128K, but I had no interest in paying the exorbitant prices these original Macs go for. The 512K on the other hand features virtually the same hardware as its slightly older cousin. It was released in September 1984, only 9 months after the original Macintosh. Other than larger RAM chips, updated ROM and a bit of logic to handle the larger memory, it is virtually undistinguishable from the 128K.

But why get the almost-the-least-powerful Mac in the first place? While I do enjoy pushing old hardware beyond its limits, the primary reason why I tinker with vintage computers is to understand how they work and, most importantly, understand how they influenced what came later. The original Macintosh was incredibly underpowered, but it had such an incredible cultural influence and long lasting legacy, so much so that I’m writing this post on a laptop carrying the same Macintosh name 39 years later. I wanted to learn more about how it all began. Andy Hertzfeld’s book was an exciting read, with many interesting anecdotes and details about the Mac’s development (these and many others are documented on Hertzfeld’s website folklore.org), some of which included examples of how its hardware and software work. Some of these examples featured limitations that I knew did not affect the later Macs like the Macintosh SE, which is another classic Mac that I own. I wanted to experience those limitations. The $80 Macintosh 512K thus became a test platform and a companion to the book written by one of the creators of the original Macintosh team. But it took a while until the Fat Mac was ready to use…

The smoke test

The important thing to note – the Mac turned on. When I first plugged it in, it booted up! But it was a bit finicky, with the screen going in and out. I actually had a friend hold it for me for a couple of weeks, and I went by his place to test it out and pick it up. After we established that it sort of works, we went to tinker with something else. Within 30 minutes we smelled smoke and feared that the house was on fire. Luckily there was no fire, but the smoke did come from the Mac – the Rifa safety capacitor burned up. The capacitor is placed in line with the mains electricity, even before it hits the power switch, so it was energized and burned up while the Macintosh was off (since we left the Mac plugged in). A quick peek inside confirmed that the Rifa was the culprit

While waiting for the Rifa replacement, I tried to figure out the reason for my intermittent image. With some experience, I now know that the cause was the solder joints on the connector between the analog and the logic board. But instead of looking up the top 3 issues with the Macintosh 128K/512K/Plus (they share the same analog board), I decided to track issues my own way… I concluded that the voltage adjustment on the analog board was partially broken internally — if I poked it (with a non conductive stick), I would get the CRT to work.

I also discovered two expansions: Dove MacSnap RAM and SCSI upgrades. The SCSI expansion worked, but the memory upgrade resulted in a sad Mac:

Analog board series of unfortunate events

I left the expansion aside for later. I also discovered that the ejection motor on the floppy drive was broken. This wasn’t too disruptive, since I have a Floppy EMU, and I could just use a stick to manually eject disks if necessary. I opened up the analog board to replace the Rifa capacitor.

The Rifa replacement went fine. But I couldn’t rest and I wanted to figure out why I had to press the R56 variable resistor to get the image to show up. So I replaced it. It didn’t help. Then I went to reflow the solder joints on the connector between analog and digital board (this was actually a good idea!). And there must have been charge left in the board… Or maybe it was static… In either case there was a spark, and I must have shorted something. From this point on I couldn’t turn on the machine. I’m going to summarize the troubleshooting steps, but if you want to read through the agonizing two weeks of troubleshooting I performed, check out my full Mastodon thread.

First I heard a clicking noise, just like in this video:

The first thing I did was test the logic board. I did this by performing a convoluted and precarious setup involving a Macintosh SE analog board, hooked to 5V, 12V, -12V and ground on the 512K logic board. I used a pair of headphones so that I could hear the chime, since the speaker for the 512k is located on the analog board, unlike the SE that just has it connected to the logic board. It beeped!

I then started replacing components, in a very chaotic and non-systematized way. Replacements in order I performed them:

• CR20, CR21, Q3, CR1, U3 and the Q8 crowbar
• Q9, a bunch of caps: C1, C2, C24, C25, C26, C29, C30, C31, C35, C36.
• CR5, C3, C43, C39, C9
• U1 (LM324N)
• U2 (74LS38)

I eventually found an actually broken component – U2 which is a 74LS38. I took it out and it failed a logic test in my EPROM programmer. It was only at this point that I realized that my tests were incorrect all along — @DosFox on Mastodon alerted me that I cannot turn on the Mac analog board without the logic board. Prompted by @DosFox‘s note about testing with the logic board, I tried to remove the monitor from the test (just leaving the tube disconnected), and actually got a good chime! Then I hooked everything up, still had good chime but no monitor output, though high voltage did go into the tube. Discharging the monitor caused a spark, which wasn’t the case for the last few days. But then I got what sounded like a bad, elongated chime:

It took me days to figure out what was happening here. I eventually realized that it was a good chime, just very long. I found this post, which suggested that the issue was caused by a bad LAG PAL 16R8. A quick scope showed that most pins seemed to be doing things, but pin 18 (labelled RESLIN on the schematic) was stuck high. Unfortunately, even though people in listservs and forums have been talking about reverse engineering this chip for twenty years, there is still no copy of it that I could program onto a GAL. It’s unfortunate, because good folks at pldarchive.co.uk have reverse engineered almost all Macintosh Plus PALs (another shoutout to @DosFox!), and Plus PALs are compatible with the 128K and 512K. Without any alternative options, I got a “for parts” Macintosh Plus board to scavenge the bad PAL. Before replacing it, I scoped pin 18 RESLIN and confirmed that it should be rather active, unlike on the LAG on my 512K board. Then I took out the bad LAG, put in a DIP 20 socket and put in the good LAG on the 512K board.

I finally got a good chime again!

Despite this success, I still didn’t get video output on the screen. I basically took a 2 week detour resolving issues I caused for myself, but I still haven’t resolved the original problem. The Mac chimed, the CRT would turn on, but I would get a horizontal line at the bottom of the screen. Playing with adjustments, I could make it go away, or make it brighter, but I couldn’t get an image. There was no vertical sync.

While troubleshooting this, I came across an issue with resistor Q7, which is placed just before the mixed video output goes into the CRT. The lead came from this thread. I was sure that was going to be it. But no, I still don’t have a raster.

Finally, I figured it out.

It was so silly. The final issue ended up being U1 LM324N, which I replaced preemptively with a replacement part earlier in the troubleshooting process. When I put the original back in, I finally got screen output. This thread helped. With the analog board fixed, I was able to test my donor Plus board (by putting its good LAG chip back in) and confirmed that even though it was sold “for parts” it was actually fully functional. I stored it for later and hopefully one day I get to source a replacement PAL, or the good reverse engineering folks will get a GAL file for it.

A lesson learned is to start using a 10 megaohm transistor when draining the CRT. The amount of fried components in this Macintosh restoration was insane. I’m pretty sure that discharging was how I fried the LAG.

Floppy eject motor

With the Mac finally working, I got started on cleaning and repairing the 800K floppy drive. It was dirty, but not the worst I’ve seen. It was actually stuck in the loaded position, even without a disk. After manually “ejecting”, I was able to read a floppy, but then the Mac couldn’t actually eject. One of the gears had actually disintegrated, making the eject motor unusable. I ordered a piece to be 3D printed, using this model. This repair also went sideways very quickly. My 3D print was too narrow (I must have selected an incompatible resin), and I ended up damaging the plastic inside the ejection motor. This spiralled to the point where I found myself trying to rebuild the motor itself, a good indicator that I took things too far, again.

I found someone on Discord and got a replacement Omron R2DG-38 ejection motor. The Mac would finally eject floppies!

Keyboard

When I got the Macintosh 512K, I didn’t have a keyboard. A friend lent me his mouse, but for a keyboard I first ended up putting together an Arduino based PS/2 adapter, by following instructions by Steve from Mac84.

This was neat and pretty easy, but it didn’t quite match the style, so I soon found a relatively affordable keyboard on eBay. I got an M0110 keyboard in pretty good shape. The =/+ keycap was missing, but the switches were in good shape. In addition to the broken +/= switch (+ missing cap), I found issues with three more keys (9, I, backspace). I combined the parts from different switches with the broken +/= switch to repair all of them. I then ordered the replacement cap and switch.

Of course, I had to clean it up as well.

When I got the missing cap I put it all together and it now looks great!

Serial adapter

I really wanted to print from my Fat Mac. A while ago I got an ImageWriter II, but it uses a 8 pin mini DIN serial cable that was standard serial on later Macs. I ended up building an adapter and using the following signal path:

Mini-DIN8 to DB25 cable > DB-25 to DB-9 converter > homemade DB-9 RS232 to Macintosh DB-9 RS422 adapter > Macintosh DB-9 printer port.

I figured out the pinout from this excellent article and the diagram below.

Original ROM

In my effort to experience the original Macintosh through the 512K, it was important to use the same ROM that shipped in January 1984. I discovered these ROMs in my Mac:

• 342-0341-C
• 342-0342-B

These were Macintosh Plus ROMs, which shipped with the Macintosh 512K enhanced (512Ke). The 512Ke had the 800K disk drive and 128K of ROM used in the Macintosh Plus. However, I suspect that my Mac received the 800K floppy and ROM upgrade later, since my model number is M0001W, which is the original 512K, while the 512Ke is model M0001E. In any case, it was time to downgrade.

It was easy to downgrade since the ROMs are available on Archive.org. I also had a pair of 27C256 EPROMs perfect for this use. I burned the 64K ROM 28BA61CE. The nomenclature here gets a bit confusing, since the ROM on the 128K and the original 512K was only 64K, while the Plus and the 512Ke had a size of 128K. So you could refer to the Plus ROM as the 128K ROM, and the 128K ROM as the 64K ROM. In any case, how do you know you’ve got the original ROM used on the 128K Mac? One way is to successfully boot the System 0.85 (pre-release) image, which can be found online.

It will boot fine on a 128K/512K ROM. The 512Ke/Plus ROM will fail like this:

Downgrading to 128K (SMC Fat Mac Switcher)

The 128K ROM helped me get closer to that original experience, but I still had 4 times as much RAM. I needed to go deeper. Luckily, just about the time I got my 512K, I came across this video by JDW:

I just love this project. It’s the SMC Fat Mac Switcher by Kay Koba. It lets you switch between 128K and 512K. It was basically made for folks doing silly experiments like me. The board is very simple. It attaches in place of a 74F253 dual 4-input multiplexer and a few resistors. It actually has those on itself in SMD form.

The board is simple, installing is not, since you need to remove the 74F253, and it has a bunch of pins tied to the ground. Desoldering ground pins on a Mac board is painful. You can guess where this story is going. I started by removing the 3 resistors and a single capacitor. That was easy. Then I moved onto the chip. Instead of snapping its legs off immediately, I tried to remove it without damaging it.

That was a mistake. About 30 minutes into trying to desolder it and failing, I snapped its pins off. But it was still incredibly difficult to get solder out of the ground pins holes. By the time I did, one pad was damaged (5) as well as one trace of pin 4 going to 26LS32. This required some bodge wires unfortunately.

The Fat Mac Switcher goes directly onto the board, but it consists of two parts that can be connected by a cable, for better placement of the switcher. You just toggle the switch to give your Mac 128K or 512K of RAM.

The 128K can really handle only a few pages of text at a time. When saved, it equals to 23KB, all the other memory is used to run the system and MacWrite. Another interesting observation on the Mac 128K memory and ROMs is that if you’re using only 128K of memory with the 512Ke/Plus ROMs, you can’t run any version of Mac paint. However, when you combine 128K with original 64KB 128/512K ROMs, it works fine. I guess it’s a matter of a few less KB of ROM getting loaded into memory.

Battery holder cleanup

The 4.5V battery in the analog board leaked at some point. It didn’t really damage anything, but it was nasty and I wanted the computer to keep the time. I cleaned up the holders in vinegar overnight, and then I got a replacement battery.

Upgrading the downgraded computer

After a while playing with my fake 128K, I decided to test out the Dove MacSnap SCSI and memory upgrades. They are not exceedingly rare, I don’t think, but they are interesting. Since the 512K wasn’t built with any upgrades in mind, these extensions literally snap onto existing chips. The memory upgrade snaps onto memory chips (and some logic), while the SCSI upgrade goes into the ROM sockets (and then the ROMs are inserted into the extension. The resulting setup is incredibly tight fitting, but with some care it does fit inside.

The SCSI extension worked fine with my DB-25 BlueSCSI, though I needed to provide external power to start it up. The memory expansion however was erroring out with a Sad Mac. While you can use error codes to identify bad chips on the 512K board, these errors cannot indicate bad chips on the expansion since the computer is not aware of the expansion board at all. I wasn’t too optimistic about fixing this, since there are 16 chips total (for an upgrade to 1MB). I got some spare memory chips and started by removing the first one. But then I got lucky! I thought I would have to get at least halfway through the 16 memory chips, but the first one did it! I replaced a 21256 with a 41256C. With the upgraded ROM, 800K drive, 1MB of memory and SCSI, this Macintosh 512K is essentially a Plus. For now I’m mostly using it in this form, since it really is more usable than a fake 128K.

Conclusion

And this wraps up my Macintosh 512K project. What started as an attempt to experience the least powerful Mac ever snowballed into one of the most challenging vintage computing projects I’ve tackled (at least until I faced the SE/30, which is an ongoing saga in itself). I made a lot of mistakes along the way, but I did learn quite a bit, not only about how little a Macintosh 128K or even a 512K could do, but also how it did certain things. I really like this machine, and I’ve grown to love its beige aesthetics. While standardization of expansion, ADB and serial ports, RAM SIMMs and other improvements that were made on the Plus and especially the SE make perfect sense, the anachronisms of the 512K are quite charming.

The post There and back again: a Macintosh 512K saga appeared first on Vlado Vince.

Note: this article was originally written for the Club de Aventuras AD magazine. You can read it in Spanish translation starting at page 68 of Issue 60 of CAAD.

Home computing in Yugoslavia, a country that no longer exists, was quite interesting in the 1980s. The geopolitically unique position of this socialist country during the Cold War produced a unique situation in terms of access to computers and technology. Yugoslavia was one of the leading countries of the nonaligned movement, a group of countries primarily from what we now call the Global South, who sought a “third way” path between the capitalist West and the communist Eastern bloc countries. This meant that the Yugoslavs had relatively relaxed access to travel to both the East and the West, and to import and export technologies to both.

While this didn’t mean that the Yugoslav could as easily obtain computers as people in Western Europe (look up the story of the Galaksija DIY computer), they still mostly used American and British machines like the Commodore 64 and the ZX Spectrum. Early in the 80s they had to find creative solutions to import them (ie. smuggle them through customs), but by 1985 such computers were being sold in Yugoslavia directly. Sinclair even partnered with a local company called Iskra to assemble ZX Spectrums in Yugoslavia. By late 1984 it was estimated that there were between 20,000 – 30,000 ZX Spectrums in Yugoslavia, almost all of which would have been brought in from abroad.

Smuggling was a national sport – one of the perks of Yugoslavia’s good relations with the West was easy access to countries like Austria and Italy, where cities close to the border (primarily Trieste in Italy and Graz in Austria) became shopping meccas for the Yugoslavs looking to purchase goods that were difficult to find locally. The catch was that while they were free to travel back and forth, they were limited to importing only 50 Deutsche Marks worth of goods per trip, a limitation that made even the cheapest microcomputers technically illegal to bring with you. Despite this, enforcement was relatively lax, as you can see in the comic strip above, published in November 1984 just as ZX Spectrums started to get sold in Yugoslavia. 

It’s primarily the experience of having to find “creative solutions” to get a computer that inspired one of the first Yugoslav adventure games – Kontrabant (“Contraband”). This game, written by a pair of Slovene students and published by the Radio Študent youth radio station, has the player travel around the country (and beyond!) to collect parts and ultimately smuggle a computer. The game was (appropriately) released for the ZX Spectrum at the time when virtually all ZX Spectrums would have been obtained in a similarly dubious fashion. 

The game sets the scene: Life is boring – the TV shows the same old movies, nothing interesting is on the radio, the news and magazines are boring. The only solution is to find a TV, a tape deck and a microcomputer. You get out of the house and find yourself in the middle of Ljubljana. A man puts a badge on you that says “Attention – Beginner Adventurer”. From this point on your adventure begins. Through the course of the game you collect items and meet characters looking for different goods. Some of the characters are famous smugglers from Slovene history who help you on your way. You travel across Yugoslavia exchanging different goods and obtaining items you’ll need to assemble your computer – you’ll need a TV and a tape deck, but you’ll also need a picture for your passport so that you can leave the country. The game is simple, but it requires a lot of contextual knowledge of both contemporary Yugoslav culture, as well as Slovene myths and historical characters. It’s also quite funny and random at times – for example, one of the items you obtain in Austria is a porno magazine, which you then exchange for some other goods. You continue collecting items until you finally assemble a computer! 

This game was quite successful, and it spawned a sequel – Kontrabant 2. Unlike the first game, which only features text, Kontrabant had some nicely drawn images. Thematically, it moves away from the “smuggle a computer” goal and instead takes the character on another tour de force journey across Yugoslavia, encountering monsters and other challenges as you try to make your way to the year 2000 and to the amazing computers of the future. Like the first game, there is a political subtext and critique of social and economic conditions in 1980s Yugoslavia. The tape also features a couple of punk rock songs by the “Kontra Band”, thematically relevant to the game.

“Yugosaurus” monster in Kontrabant 2

Another popular Yugoslav adventure game was Vruće ljetovanje (Hot Summer) where you need to get your family over to the coast for vacation. You keep running into increasingly difficult issues and challenges. You start by waking up and you need to get some breakfast. This simple task teaches you the game mechanics, because the game is timed and you need to eat and rest (or you will die!). Your character, a father called Srećko, has to get money and food for the vacation, then when he tries to book a trip he finds out that all the spots are taken. Luckily, he bribes the travel agency clerk and books the vacation.

The game is interspersed with tasks and situations typical of the Yugoslav working class life in the 80s, the challenges caused by rapid inflation, low level corruption at every step and other daily difficulties. When you finally get to the coast things get really interesting. A certain Maja leaves a message for you to meet her at the nudist beach. She had her ring stolen and asks for your help retrieving it – turns out a group of gamblers is staying at the hotel with a load of stolen goods. You decide to help Maja, but you also need to make sure your wife doesn’t think you’re cheating. Through a series of twists and turns, you manage to recover Maja’s ring and other jewelery, and as a reward you get another week on the coast with your family.

Due to complex mechanics with time, hunger and rest variables, Vruće ljetovanje was arguably the most difficult Yugoslav adventure game. Additionally, the syntax could get incredibly complex, with sentences of up to 10 words! Very few players would have finished this game, but its highest value is in its historical and local context. Players familiar with daily life in Yugoslavia will recognize the challenges of the time, and local references to locations, companies and pop culture references are integrated with the gameplay seamlessly. Additionally, the illustrations used in the game are very well done. They were drawn by Igor Kordej, a comic book artist who would later on go to produce work for both Marvel Comics and Dark Horse Comics. 

In addition to the Kontrabant series and Vruće ljetovanje, there were maybe a few dozen other domestically produced Yugoslav adventure games. Like with other software and games at the time, due to easy access to (most often pirated) foreign titles the domestic production was relatively limited. A lot of homebrew adventures were simple crime mysteries, with noir elements. While most of these other games were pretty simple, a few more interesting ones include Bajke (“Tales”), based on classic Slovenian children’s tales, XIV (“14th”) series about adventures in Belgrade’s XIV Gymnasium and Na Balkanu ništa novo (“Nothing new in the Balkans”). This last one is particularly interesting, as it’s one of the final Yugoslav games before the break up of the country. It explicitly deals with the then brewing political conflict, featuring many real life characters, some of whom would become leading figures of the breakup of Yugoslavia.

Domestic adventure games in Yugoslavia represented a relatively minor part of computer gaming experience, but they were a fascinating platform for experimenting with both alternative forms of narrative, but also as another avenue for political expression. Both of the featured games give us great insight into daily life of Yugoslavs in the 1980s in a way that other games produced locally do not – it’s specifically through the focus on the narrative that the local context is centered.

The post Yugoslav adventure games appeared first on Vlado Vince.

Introduction

In early June I went to the VCF Swap Meet in Wall, NJ and picked up a supposedly functional TRS-80 Model I. I haven’t been very interested in TRS machines (other than owning a CoCo 2 for a moment) until I watched a video by Dejan Ristanović, one of the creators of the Galaksija project, who shared some less known facts about the history of the Galaksija and its connection to the TRS-80. In this video, Dejan mentioned how the TRS-80 influenced Voja Antonić, Galaksija’s designer. Voja’s first original computer design in 1981 was actually based on the TRS-80, but he significantly cut down the number of ICs, upgraded memory to 48K and implemented a number of other improvements. Voja did not have a name for this computer, but Dejan called it “TRK+ TRK-“. Additionally, TRS-80 Level I BASIC, based on the public domain Palo Alto Tiny BASIC by Li-Chen Wang, influenced Galaksija’s BASIC. As Dejan notes in his video, this influence was miscredited to Microsoft for decades due to the wrong assumption that Microsoft provided Level I BASIC for the TRS-80 (they did write Level II and Level III). Even the keyboard on the Galaksija is heavily influenced by the TRS-80, as you can see in the left and right side location of the arrow keys.

With all of this, I decided to get a TRS-80 Model I to better understand the prehistory of the Galaksija. I started looking for one online, but the few rare listings I came across went for hundreds of dollars, which was way more than I was ready to spend. Back in May, I found one at VCF East consignment for $100, but since I was exhibiting, I had very little time to purchase things and I missed out on it. I decided to wait for the VCF Swap Meet in June and figured there would be someone selling one.

I was right! There was a vendor selling a couple of TRS-80 Model Is, with expansion interfaces and monitors. Since I took the train + folding bike to the swap meet, I had no way to transport the monitor (nor space at home to store it), so I made a deal to buy just the computer. The price of $95 was a bit higher than I was planning to spend, but the seller claimed it was fully functional, and the TRS-80 was the one thing I wanted to come home with.

The TRS-80 was in “not great, not terrible” shape. Pretty dirty, with some paint scrubbed, but generally it was in decent enough physical condition. I started by making a composite cable for it (since I didn’t take the monitor). The TRS-80 Model I uses a DIN-5 connector with a 5V line that goes to the TRS-80 monitor, but otherwise it’s a regular composite connection. Once I built the video cable, I tried to turn the computer on. First there was a good sign — the indicator LED lit up! Unfortunately, I didn’t get video output. There was no output to my composite to HDMI adapter, so I grabbed the only CRT monitor I own — the Apple Monitor IIc. I got very faint output without horizontal sync. It was time to open the case.

As soon as I opened the case I noticed 3 things

• there was a small daughter board with the Z80 CPU
• the keyboard connector had disintegrated, and a previous owner soldered bodge wires on top of it
• there was significant corrosion on chip Z6 (74C04)

Video circuit repair

I started by dealing with the corroded chip. A quick search of common TRS-80 issues confirmed that Z6 is a part of the video circuit, and a very common component to cause issues. Getting the old chip out was difficult, and I had to cut the legs off of it to remove it. I also had to be very careful not to lift any pads, as the PCB is in rough shape. I put a socket in its place and added two bodge wires to fix the two connections that were broken by corrosion.

At this point I made a mistake that would take me a few days to figure out — I replaced the 74C04 chip with a 74LS04. Using an LS chip made sense to me since most of the other logic chips on this board were LS series. I later realized that I needed to put a CMOS chip there instead.

Anitek Supermem 256/512k memory upgrade

Even with this adjustment, I still had no video output. I decided to remove the strange daughter board. It had a bunch of wires soldered onto the board, including two going to 74LS157. At this point I noticed bodge wires going from the dynamic memory chips, connecting their PIN 9 (address line A7), to 74LS157 as well. Even though I didn’t yet know what this expansion actually was, I figured it had something to do with memory. I then noticed that the memory chips were 41256 — way bigger than the 4116s which would normally be the largest chips on this board to give it 16KB of RAM.

In the meantime, I tried to find out more about the daughter board expansion. I posted about it in the TRS-80 Model 1-4/4P Facebook group and soon found out that it was the Anitek Supermem 256/512k memory upgrade from the late 80s. The board with the PALs is a bank switchers and logic chips to handle the 256 cycle refresh memory chips.There’s more info in this VCF forum thread. I decided to take it out for now and perhaps try to get it working later.

Video circuit repair continued

I fully removed the daughter board, and moved the Z80 back onto the board. I tried video output again, with no luck. At this point I started looking around the dynamic memory chips again, and I noticed a number of modifications — cut traces, bodge wires. I pulled up the datasheets for 4116 and 41256 RAM which confirmed what I expected — there are a few differences between them, primarily in terms of power requirements. 4116 needs -5V and 12V in addition to 5V. 41256 only needs 5V. The modifications gave 5V on pin 8, which is 12V on 4116. -5V was completely removed (originally pin 1), while the pin 9 (originally 5V on 4116) was connected to 74LS157, presumably to be able to use address line A7.

I also noticed that a number of passive components were missing from the board, including R19, C3, C4 and CR2, which according to the schematic are a part of the -5V circuit.

Even though the 41256 RAM wasn’t using the -5V power, I figured it wouldn’t hurt to put those components back. I did not have an IN5231 compatible diode, but I did put C1, C4, C3 and R19 in place. Once I did that, I finally got something on the screen!

I got a pretty picturesque error – the computer was looping with MEMORY SIZE ?SN ERROR. This had me stumped for a few minutes, because I thought it was a memory error. However, this article helped me understand that the error was simply due to the keyboard being disconnected.

Keyboard connector repair

I had planned to tackle the keyboard later. The bodge wires the previous owner put in place were falling apart, so I decided to completely replace this with a ribbon cable. I got the idea from the aforementioned article. I’ve made some ribbon cables before, so this was a relatively simple job. I soldered angled pins aimed inward. This is important because otherwise the ribbon cable wouldn’t have fit in the case

Once I put the ribbon cable together, the computer finally booted successfully!

Memory restoration

I was able to issue a few commands, but very unreliably. I kept getting errors like ?OM ERROR. PRINT MEM showed only 81 KB available. 41256 RAM was working, but with an extremely limited amount of memory. I went back to the data sheets for the 4116 and 41256, trying to get the 41256 to basically behave as the 4116. Once I tied A7, which had been going to pin 4 of 74LS157 and the daughterboard, to ground, I finally had access to all 16K of usable RAM (max for the TRS-80 Model I without expansions).

XRX modification and cassette input

The computer was functional, but the screen output was quite wobbly. The next step was testing cassette input. I made another cable, this time making a 3.5mm to DIN-5 cable. I got the TRS-80 tool to convert .cas files to wav and converted a few programs. Before attempting tape load, I made sure that the XRX modification was on the board. This modification is a small PCB with two logic chips on the bottom of the board, meant to improve loading from tape. Initially I took it out because I didn’t know if it interfered with anything. It probably made no difference as it was an official fix/upgrade that Radio Shack would add to existing computers due to a bug in Level II BASIC that introduced loading issues. Mine is likely version XRX-III of the XRX. XRX was made redundant by version 1.3 of the Level II ROM.

Once the XRX modification was back on the board, I successfully loaded Star Trek!

Keyboard cleaning and repair

I still had to deal with the image wobble, but I took a break to wash the keys. I also had to repair a few keys that weren’t responding well. Adding some fresh solder the key switch connections seemed to fix the issues reliably.

Image wobble repair

The last thing to fix was the image wobble. It was very visible when looking at some kind of graphic like the frame below. Following suggestions online, I started replacing some passive components like electrolytic caps and resistors that are a part of the video circuit.

I had no luck, not even when I replaced the 4 variable resistors in the video circuit (R20, R21 and R5, R6). I then noticed the difference in the video signal HSYNC when comparing its wobbly signal to the good one from an Apple //e:

It looked like my H sync pulse was closer to 20μs vs 4μs cited in this article, but it was “jitter on H sync pulse = wobbly picture” that caught my attention because I definitely had jitter on that signal.

TRS-80 power supply issue

I had also tested the power supply and noticed some issues — I got 17.2V DC and 19.2V AC. Per the TRS-80 Microcomputer Technical Reference Guide AC should be 14VAC and DC should be 19.8 VDC. I even opened the power supply and tried changing the single diode. This had no effect and I accidentally blew the 40 year old fuse so I had to replace it with a new one. Opening the case was extremely difficult and it required me to cut the plastic. When I put it back together I used electrical tape to close it back up.

Wobbly video repaired

After replacing at least a dozen passive components around the video circuit, I finally took out my scope and started looking at the signal, starting from the composite video connector, until I found the source of the noise which I presumed was the cause of the wobble. I ended up finding something wrong on pin 10 on Z6 74LS04.

At this point I remembered that I replaced 74C04 with 74LS04. As soon as I put a CMOS version in its place (74HC04), the wobble went away for the most part.

Conclusion

The image is still not 100% fixed (for example, I still can’t reliably use the composite to HDMI adapter) but it is fully usable with the Apple Monitor IIc. I will likely need to obtain/build a new power supply to know whether that’s what’s been causing any of my issues. This post may expand in the future, but for now it covers most information about my repair of this TRS-80 Model I. As most of my repairs and restorations, this post is based on a thread on Mastodon.

The post TRS-80 Model I restoration appeared first on Vlado Vince.

I started by opening up the iPaq. It was relatively easy, but it required a TR6 screwdriver. Once I took out the 4 screws, it was easy enough to remove the case. There are detailed instructions here if you need help — make sure not to lose the battery door cover and the stylus latch. Once I got the the battery, I confirmed it’s a Danionics P/N 167648 DLP 305590, 3.7V battery. It was soldered onto a flex ribbon cable with the connector that goes to the PCB. After a quick online search, I confirmed that there were no compatible replacements available anywhere anymore.

I started by desoldering the ribbon cable from the battery, and started looking for a suitable replacement. My first candidate was a Google Pixel 4A battery I had. I soldered it on and verified that it can hold a charge.

Unfortunately, I soon found out that even though it was the right width and length, it was too high (thick). When I tried to screw the back cover back on, the battery started to press against the PCB, pushing the screen and risking damage. It was a good proof of concept, but not yet ready for use. I started looking for a replacement.

I confirmed that the height (thickness) of the original battery was 4.8mm. I started to search for a replacement online and was surprised how difficult it is to find a battery so thin. I came across HXJNLDC DC 3.7V 1300mAh 503759 which was supposed to be 5mm thick. I figured that the 0.2mm difference might work, so I ordered it. Unfortunately, that wasn’t the case. Once I got this battery, I ran into a similar issue with it pressing against the screen — I had to look for yet another replacement. I finally came across KZYTUE 3.7V 1100mAh 454950, a battery with slightly less capacity, but supposedly only 4.5mm of thickness.

KZYTUE 454950 next to the iPaq and the original P/N 167648 battery

I went ahead and cut off the connector from the battery so I could solder it to the flex connector.

Luckily, this time around the battery fit perfectly and I was able to reinstall the back cover with no issues. After testing the battery for a while it’s pretty reliable. So, if you’re looking to replace a battery on one of these 3600 series Compaq iPaqs, I recommend KZYTUE 3.7V 1100mAh 454950.

The post Sourcing a replacement iPaq 3600-series battery appeared first on Vlado Vince.

Each character consists of 16 bytes which correspond to the number of bits that are turned off for each visible pixel (pixel = 0), going from right to left. In the case of the letter Š outlined above, the first row equals 10101111= AF in hex.

Characters are stored in the order below:

This means that the letter Š is the 30th symbol in the ROM (counting from 0). The 16 bytes for each character are 128 bytes apart each, because the Galaksija knows 128 characters total. Characters start at address 0000 0000 with the top row of the left side of the ready symbol:

Since the top row is empty, the value of byte 0000 0000 in the character generator ROM is FF. We can verify this by dumping the first symbol (left side of the ready symbol) via the chargendump tool:

We see that the first two rows are empty, but the third row has one pixel lit up, so we expect that the third byte will be: 10111111 = BF. To verify, we’ll search the ROM for byte 256 (because it’s the third byte and there are 128 bytes between each character’s row) = 100 in hex:

Checking the ROM, we see that byte 0000 0100 (byte 256 decimal) is indeed BF.

We can now use the 8×16 matrix to draw new characters. Let’s say we want to draw a star instead of the Elektronika inženjering or Mipro logo for our READY prompt. Since the READY sign consists of two symbols, we’ll need to edit both symbols 0 and 39.

Symbol 0 will be:

0 FF
80 FF
100 FF
180 7F
200 BF
280 C1
300 FB
380 F7
400 6F
480 6F
500 B7
580 D7
600 E7
680 FF
700 FF
780 FF

Symbol 39 will be:

27 FF
A7 FF
127 FE
1A7 FD
227 FB
2A7 87
327 DF
3A7 EF
427 F6
4A7 F6
527 ED
5A7 EB
627 E7
6A7 FF
727 FF
7A7 FF

Final result:

This should be enough information necessary to create new characters and properly change them in the ROM. To learn more about how the Galaksija generates characters, read this article by Tomaž Šolc.

The post Designing custom characters for the Galaksija appeared first on Vlado Vince.

If you’re looking for the replacement SMK style switches, I got mine on eBay from this seller.

When building my first Galaksija, I bought some Cherry MX compatible Gateron keyswitches. I soon found out that not only did my board not have mounting holes, the connectors were also too far from each other to work with these standard switches. I ended up soldering some extra wire onto these switches in order to get my Galaksija working, but the keyboard was horribly unstable and very unpleasant to use. The only thing I knew about the original key switches was that they were meant to be IEVT TY or TX switches — because they were listed as such in the original parts list.

No other information was provided that would explain what standard these key switches were. Back in 1984, key switches were among the parts that the Računari u vašoj kući magazine offered to folks as part of their kit (along with the PCB, keycaps and the keyboard plate). The key switches were manufactured by Inštitut za elektroniko in vakuumsko tehniko (IEVT) from Ljubljana. That was the extent of available information I had. I completed the Galaksija with my improvised Gateron switches and hadn’t thought much about them for a while. Finally, a few weeks ago someone provided me with a scan of the IEVT catalog featuring TY switches, and this restarted my research.

Last year I obtained a pair of Orao computers, which also use IEVT TY switches. Since I still had no idea what IEVT TY switches actually looked like, I decided to take one out and test it on my single layer board Galaksija. I quickly confirmed that these were indeed compatible, even though these particular switches had 10% angle and thus wouldn’t be usable without modification.

Back of the IEVT TY switch

IEVT TY switch on a single board Galaksija (10% angle switch from the Orao)

Most importantly, I realized that the switch format I was looking for, with two connector holes in the middle of the switch, was just incorrect. Ever since I made my Galaksija, I’ve been looking for a switch that fit this footprint:

But this was a mistake with the silkscreen I got from the gerber on специалист-пк.рф/Galaksija, which was based on a gerber made by demerzel on forum.benchmark.rs back in 2013. On their gerber, the keyboard outline was pushed slightly down, making it seem like these keys switches looked different than they actually do.

With this confirmation (and A LOT of internet archaeology tracking the gerber files and when the original mix up was made with the silkscreen), I was able to continue the search for what IEVT TY switches actually were. A friend pointed me to Tipro SMK Style deskthority page, and this was the final piece of the puzzle. These key switches were virtually identical to my IEVT TY switches from the Orao, and due to branding in both the aforementioned catalog and my own switches, we were able to confirm that Tipro was the same factory that produced the TY switches — Slovenian company Tipro was founded in 1985 by engineers from IEVT (Institute for electronics and vacuum technology). Tipro produced a lot of products developed at the IEVT, and it was privatized in 1991 along with other state owned companies in Slovenia. They are still in business.

But back to TY switches — thanks to the Deskthority page, we were able to confirm that they were SMK J-M0404 series style of switches, the same kind that was used on some Apple //es and BBC Micros. I still don’t know if these were unauthorized clones, or if they were made under license from SMK, but they are certainly the same style. Luckily, I was able to verify this final test — I have an Apple II with this exact same type of switches! I took out one switch and was immediately able to verify that it fit the Galaksija perfectly!

This test however verified one more thing — the gerber file that had the mistake on the silkscreen also didn’t have wide enough connector holes in the drill file, so in order to fit SMK style switches, the holes had to be widened. But this was a minor annoyance now that the mistery of the key switches was solved. I used a drill to widen the connector holes (see below for a photo).

The final step was sourcing replacements. I decided not to search for original IEVT TYs — it was unlikely that I would find any new old stock, and I had no interest in scavenging another vintage Yugoslav machine like an Orao just so I could get switches for my replica single board Galaksija.

I did look for another source of SMK J-M0404 replacements for Apple //es, but they ended up being rare and expensive. Finally, thanks to some kind help, I found a source on eBay for new SMK style switches, and they even came with transparent relegendable caps!

Thanks to this, I’ve been able to use all the correct symbols on all the keys — I used Tomaž Šolc’s pdf file (inside this zip) to print out all the symbols. The shipping was a bit expensive, but I got them pretty fast. Now it was time to add them to my Galaksija! Technically I’m still missing the plate, but these key switches are already so much more stable than my original solution that I’ve decided not to wait until I can replicate a plate (although I do plan to make one some day)

Keyboard plate instructions in Issue 1 of Računari u vašoj kući

I’m pretty happy with the final result! I used a Cherry MX compatible spacebar (which fit ok on the SMK style switches), but I still need to find a good enter key, since it needs to fit on top of two switches. My space bar also need a stabilizer.

I’m very inexperienced when it comes to keyboards and key switches, so this turned into quite an odyssey, but I’m pretty happy with how it turned out. I hope this helps you if you’re trying to track down some of these ancient key switches and understand what’s what. I leave you with some extra pics of original IEVT TYs, made in Yugoslavia!

The post Yugoslav IEVT TY key switches – SMK style switches in disguise appeared first on Vlado Vince.

Last edited: 27 May 2024.

Introduction

Galaksija is a Z80A based microcomputer designed in 1983 in Belgrade (Yugoslavia, now in Serbia) by Voja Antonić. I’ve been documenting, exhibiting and speaking about the Galaksija ever since I assembled my first one in 2020.

If this is your first time hearing about the Galaksija, I recommend you check out my three posts telling the story of the Galaksija:

• Building A Galaksija, Part I: The Tech
• Building A Galaksija, Part II: The Origin Story 
• Building A Galaksija, Part III: Audio Cassettes, Magazine Listings And Radio Waves

This post includes many resources and references if you are building one of your own, or if you’re just interested in learning more about this fascinating machine. Many of these resources are also featured on my Galaksija github repo.

Sections:

• Board design and gerber files
• Other Galaksija designs
• Galaksija boards to order
• Bill of materials (parts list)
• Instructions
• Schematic
• Roms
• Video output
• Power supply
• Software and tape loading
• Emulators
• Memory map
• Case
• Key switches
• Keyboard plate
• Expansions
• Common problems
• Galaksija builds
• Galaksija Plus
• Additional resources

Board design and gerber files

The single layer PCB design of the Galaksija was originally published in 1984 Računari u vašoj kući magazine on page 56 (page 8 of the pdf). While most folks ordered their boards from the magazine, a complete board could be made from the copy included in the magazine. While you can still make your Galaksija using this same design (like I did!), there is now a double layer design gerber published by Voja Antonić, Galaksija’s creator. However, there are many other single layer gerbers going around the internet. I will summarize the different versions below

1983 single layer board

The original board was published in Računari u vašoj kući magazine in January 1984. While most folks ordered these boards from the magazine, the print could be used to manufacture a board yourself. The print was in 1:2 ratio. It was a single layer board with 119 jumpers, and it was compatible with SMK style key switches.

2020 double layer board

Voja Antonić, Galaksija’s designer, released an updated double layer PCB back in 2020. The 2020 double layer PCB includes all the connections that had to be manually made with jumpers on the original single layer PCB. It also includes updated key switch connectors and drill holes compatible with Cherry MX switches, and a new 5nF ceramic capacitor added between RFSH and GND on Z80A (alternative values like 3nF might work better, see below for common problems). This solves the problem of modified timings on the CMOS version of Z80A. You can download the double-layer PCB from Hackaday.

Single layer board gerber versions

Demerzel gerber

Gerber file of the original single layer board was originally published by demerzel on forum.benchmark.rs who made it by scanning the original board design from Računari u vašoj kući. You can get that gerber here. Note that since this is a single layer board, you will have to solder 119 jumpers, and the connector and drill hole locations are not compatible with modern key switches, so you will have to source older switches (SMK J-M0404 compatible), and possibly drill the key connector holes to widen them (see the key switches section for details). This gerber is a great reproduction of the original board design, as you can see in this comparison below where the gerber is outlined in red and purple, over the scan from the magazine in black:

Demerzel’s gerber was later repackaged (possibly with some minor edits) and made available at spetsialist-mx.ru. That URL is no longer active, but a mirror can be found at http://www.специалист-пк.рф/Galaksija. The gerber is pretty much the same as the demerzel version, including the connectors for old style SMK key switches. You can download it here.

Galaksija tlm 092013 (single layer gerber with MX switches)

Demerzel at the forum.benchmark.rs put together a version of the single board gerber with MX compatible keyboard connectors. The top part of the board is the same as the demerzel/spetsialist-mx.ru version, but the bottom (keyboard) has been changed to fit MX key switches. Note: this gerber does not include drill holes for the MX key switch mount holes, so you will either have to add the holes to the gerber or drill them. You can download it here.

Other Galaksija designs

There are other Galaksija compatible designs that have been made by other folks over the years. Here are a few that I’m aware of:

• Galaksija 2024 – a 40th anniversary Galaksija designed by Voja Antonić, featuring compact design, an EEPROM for storage and a UART for communication
• Nova Galaksija (“New Galaksija”) – Zoran Bošković’s Galaksija design from 2004 that doesn’t include a built in keyboard and fits the whole design on a small PCB. Josip Perušanec designed additional improvements for this Galaksija including better graphics, IDE interface and sound chip. Schematics are available on the archived spetsialist-mx.ru.
• Galaksija Plus – Galaksija Plus was an improved Galaksija design by Nenad Dunjić, Milan Tadić and Nenad Balint in 1985, it included popular expansion projects like the expanded memory, sound support and improved graphics. Folks at spetsialist-mx.ru designed a clone of the Galaksija Plus. The design files are unfortunately no longer available, but luckily Retro Bertie on YouTube uploaded them to Google Drive. Retro Bertie has a great video about building a Galaksija Plus clone. Fixes cited in this video were developed by Goran Stanković. More information about the Galaksija Plus is available in the Galaksija Plus section below
• CMOS Galaksija – Tomaž Šolc’s remake of the original circuit using modern 74HC-series chips. Tomaž wrote his Master Thesis about this project
• µGalaksija – an FPGA Galaksija designed by Dušan Grujić i Dragan Toroman
• Galaksija FPGA for ZXUno – FPGA Galaksija implementation for the ZXUno

Galaksija boards to order

There are currently (2023) two webshops with Galaksija boards available to order:

• galaksija.org.rs, associated with the Galaksija documentary project, has Voja Antonić’s 2020 double layer boards available for €35
• Apulia Retrocomputing has single layer boards available for €29.90

Bill of materials (parts list)

I created a spreadsheet with all the parts necessary to build a Galaksija, with links and prices. This spreadsheet includes separate BOMs for the single layer and the 2020 double layer PCB designs.

Note: this file is not actively maintained and you should check availability before ordering.

Instructions

Instructions for building the Galaksija computer were first published in the January 1984 special edition of the Yugoslav popular science magazine Galaksija, titled Računari u vašoj kući (Computers in your home). These instructions are still valid and they are the best resource if you want to build a Galaksija yourself. This translation was published in PoC——GTFO, Volume 2 published by No Starch Press in 2018. This translation is kindly provided by Travis Goodspeed and Aleks Nikolić. 

• Original instructions from Računari u vašoj kući
• PDF Instructions English Translation

Schematic

The original instructions in Računari u vašoj kući included the full schematic for building a Galaksija. While you can find this on page 53 (page 5 of the PDF) of the untranslated original guide, the English translation does not include the schematic. I’ve included both the original scan from the magazine, and the modern recreation by “fifan” from spetsialist-mx.ru:

• original schematic from Računari u vašoj kući
• schematic from spetsialist-mx.ru

Note: the schematic from Računari has the resistors R9 (62 ohm) and R10 (50 ohm) flipped. I’ve corrected the version attached here to show the correct connections. You can find the unedited schematic with the error in the Računari guide. Big thanks to Walter Miraglia for spotting this!

Roms

There were 3 roms for the original Galaksija

• 2 x 2732 4KB EPROMs: ROM A (main functions and BASIC) and ROM B (advanced math functions, released in 1984)
• 1 x 2716 2KB EPROM: character generator rom

Later on at least two additional roms were released, mainly for use with Galaksija Plus, but still compatible with the original Galaksija: ROM C and ROM D. ROM C is available, but I haven’t been able to find copies of the ROM D. More information about the Galaksija Plus is available in the Galaksija Plus section below.

In terms of flashing the ROMs, 2732 EPROMs are somewhat difficult to burn using cheap EPROM programmers like the TL866II Plus. You may have issues with the programming voltage for the 2732A EPROMs which require 21V when programming (there are some simple mods that will let you use 21V). 2732Bs are easier, since they only use 12V when programming.

However, I’ve also ran into some problems with the programming pulse required by the 2732 (50ms) when using TL866II Plus. Because of this, the flashing will only write a few bytes at a time. While I’ve previously been able to get around this by writing a script that flashes the ROM about 1000 times (check my original post about my experience building a Galaksija), there is a much easier way to do it by connecting pins 18 (/CE) and ground (pin 12) while the EPROM is in the TL866II programmer. Once you make this connection, the flash will succeed in a single go.

Your mileage may vary. Like always when using old 2732 and 2716 EPROMs, I recommend getting a few extra ones just in case.

ROM A

• ROM A version 28 – original ROM A before ROM B was available
• ROM A version 29 – updated ROM A with automatic loading of ROM B

• the listing of the original ROM A version 28 with Voja Antonić’s comments, published as GALA005052. Kindly provided by Paolo Gigli
• Galaksija ROM A dissasembly by Tomaž Šolc (in English)
• the original Galaksija software guide published in Računari u vašoj kući Issue 1, covering ROM A functions

ROM B

• the original ROM B
• ROM B with reverted screen position fix. The original ROM B moved the image position to the right. This worked well with TVs, but may not fit well on a monitor.

• the original ROM B listing (complete)
• the original ROM B guide published in Računari u vašoj kući Issue 3.

Character generator ROM

• Mipro character rom
• Elektronika inženjering character rom

While the character ROM hasn’t changed after release, different computer versions came with different logos. A number of companies were involved in production and distribution of Galaksija parts, but three companies actively participated in Galaksija branding:

• Elektronika inženjering
• Mipro (Voja Antonić’s company)
• Zavod za udžbenike i nastavna sredstva

Elektronika inženjering and Mipro featured their respective logos at the Galaksija READY prompt. I haven’t been able to confirm what determined which character ROMs got which logo. Per Dejan Ristanović, all the EPROMs sent to Računari u vašoj kući offices to be programmed by Voja Antonić were programmed with the Mipro logo, while the factory made Galaksijas shipped with the Elektronika inženjering logo. Most character ROMs available online feature the recognizable Elektronika inženjering logo, but I generated one with Mipro’s logo as well for historical reasons. The first to document this online was Tomaž Šolc who also published the comparison below:

Mipro logo

Elektronika inženjering logo

Read this article for more information on generating your own logo and creating new characters for the Galaksija.

ROM C

• ROM C
• ROM C listing (kindly provided by Hindicki Ferenc)
• a scan of the article about ROM C in Svet kompjutera

ROM D

• a scan of the article about ROM D in Svet kompjutera

Video output

Galaksija was designed with both composite and RF video output. While you can place an RF modulator on your Galaksija and use it with an old TV, you will likely only use composite output. Note that Galaksija uses PAL timings since it was designed for use in Europe! If you don’t have a PAL TV or monitor, you can use a composite to HDMI converter, but note that many cheap ones you can get on Amazon produce an unstable PAL image. This is due to the use of a 6.114Mhz oscillator which produced an image slightly out of line with the PAL standard. I’ve had a great experience with a Retrotink 2X-MINI (pictured above). If you have an NTSC monitor with composite input, you can also use a PAL to NTSC converter, or you can try using it without an adapter, since Galaksija outputs a black and white image.

Power supply

Galaksija uses a 5V power supply, with optional 12V needed for an RF modulator if you decide to include one. The original design and instructions included a schematic for building your own PSU with both 5V and 12V, assuming 220V input (since Galaksija was designed for use in Europe). The power supply design and instructions are included in the translated Galaksija instructions.

Since we have cheap and reliable 5V power adapters now, there is probably no need to build this yourself. I’ve used off the shelf 5V 2A power adapters like this one.

Software and emulation

Galaksija used cassette tape to store programs. To learn more about how the Galaksija uses tapes, read this post by Tomaž Šolc.

The best online resource for Galaksija programs and games is retrospec.sgn.net. Most programs are available as gtp files which you can load into emulators, or convert to wav for loading on a real Galaksija.

Gtp files are supported in MAME. To convert WAV files saved from a Galaksija to gtp files for storage, see my dump2gtp tool on GitHub.

Miodrag Milanović’s galtape utility is the best tool to convert WAV files to gtp files.

Loading tape programs on the Galaksija

To convert to wav files for use with a real Galaksija, use gtp2wav available as part of Tomaž Šolc’s Galaksija development tools. Once you have converted programs to WAV, you’ll need to connect your computer or another audio playing device to the Galaksija via audio cable. On the 2020 double layer board, the 3.5mm jack is connected to the right channel for loading (labelled KU) and the left channel for saving (labelled KI) to tape. Check the common problems section for help getting the tape interface working.

When you’re ready to load your program, type OLD

Galaksija will turn off the screen while the tape is loading and you won’t have any feedback until it has been loaded. In case of verification problems, you will get the WHAT? error message. If the tape loads correctly you will get the READY prompt. If the tape doesn’t load at all, Galaksija will stay on the empty screen until it is reset.

Tape commands:

OLD – loads program from tape

OLD? – verifies the program from tape against the program in memory, useful for verifying successful save to tape

SAVE – saves program from memory to tape

Emulators

• Galaksija is well supported in MAME. You can use this zip file with the 3 roms in MAME.
• GalaxyWin emulator for Windows.
• Galaxy – DOS emulator for the Galaksija. Download zip from here
• Galaksija.net is a browser emulator running the Galaksija Plus version of Galaksija written by Dragoljub B. Obradović
• Jsgalmin – Both Windows and browser emulator with Galaksija Plus support, written by Hindicki Ferenc. The emulator is included in a large zip file with a huge amount of Galaksija materials that can be browsed as an offline webpage (start by opening index.html)
• Galaksija – for Sinclair ZX Spectrum 128k
• Galaksija Emulator – for SAM Coupé

More information about Galaksija software is available in my Building a Galaksija Part III blog post.

Memory map

• $0000…$0FFF — ROM A – 4 KB contains bootstrap, core control and Galaksija BASIC interpreter code
• $1000…$1FFF — ROM B – 4 KB (optional) – additional Galaksija BASIC commands, assembler, machine code monitor, etc.
• $2000…$27FF — keyboard and latch
• $2800…$2FFF — RAM C: 2 KB ($2800…$2BFF – Video RAM)
• $3000…$37FF — RAM D: 2 KB
• $3800…$3FFF — RAM E: 2 KB
• $4000…$FFFF — unused, for extensions

Case

The original Galaksija instructions in Računari u vašoj kući suggests using plain PCB material (fiberglass) to make the case. I designed a case in SketchUp that follows this design closely, but which is ready to be 3D printed.

• the design assumes a plexiglass base with standoff screws (4)
• 2 LED holes
• 2 reset button holes (1 NMI, 1 full reset)
• 1 power jack
• 1 power switch
• 1 3.5mm jack

I strongly suggest you measure your Galaksija, especially the keyboard, and make adjustments for I/O ports. Overall dimensions should be compatible with all original PCB designs.

• SketchUp file
• STL file for 3D printing
• Additional photos

Key switches

Original single layer PCB

More information about key switches for the single layer Galaksija is available in my post about IEVT TY switches.

Back in 1984, key switches were among the parts that the Računari u vašoj kući magazine offered to folks as part of their kit (along with the PCB, keycaps and the keyboard plate). The key switches were manufactured by Inštitut za elektroniko in vakuumsko tehniko (IEVT) and the type used was  IEVT TY (the same type was used in the Yugoslav Orao computer). This key switch is a version of the SMK J-M0404 which was used in some Apple //es and other popular microcomputers in the 80s. Slovenian company Tipro took over production of these switches after 1985. Tipro was privatized in 1991 when many publicly owned companies in the former Yugoslav states were privatized. Tipro’s TY switches are documented on Deskthority as Tipro SMK Style.

IEVT TY switches are extremely rare, and even SMK J-M0404 and other compatible SMK switches are difficult to find. One option is finding another computer or keyboard with SMK switches, or you can alternatively search eBay for folks selling SMM J-M0404 replacements for Apple //es, but they tend to be rather expensive per switch. Alternatively, you can also get modern SMK style replacements like these ones. You will need 54 key switches. It’s also recommended to use a keyboard plate with SMK switches, like in the original design – check the keyboard plate section below for making your own.

Important note: if you are using a single layer board based on the demerzel/spetsialist-mx.ru gerber, you will likely have to drill the key switch connector holes to widen them enough to fit the SMK J-M0404 connectors. Also note that the connectors should be mounted on the bottom of the keyswitch (not in the middle, as indicated by the outline).

2020 two layer PCB

The 2020 double layer PCB is Cherry MX compatible with PCB mountable key switches, so a wide variety of switches and caps are available. In my BOM, I recommend Cherry MX Blue Switches PCB Mount and ABS ADA Keycap Set – Film Never Dies caps. You will need 53 key switches (1 less than the single board design because the double layer board uses a single switch for the enter key).

Keyboard plate

Since the SMK key switch type is plate mounted, it requires a plate which was provided as part of the kit, but plate design was also published in Računari u vašoj kući. I designed and manufactured a 3D print of the keyboard plate, which you can download from my GitHub as both a Sketchup or STL file for printing.

Expansions

There were at least a dozen expansions for the Galaksija back in the 80s, mostly homebrew, and many of which were developed by Voja Antonić himself. In 1985 the Galaksija Plus project implemented many of those expansions (like the upgraded memory, sound, improved graphics etc.) into a new board design. Below are a few expansion projects that we have designs for:

• Centronics (parallel) interface (article by Voja Antonić)
  • Design published in October 1984 issue of Svet kompjutera
• EPROM programmer (article byJova Regasek)
  • Design published in Issue 2 of Računari u vašoj kući
• Memory expansion (article byJova Regasek)
  • Design published in Issue 2 of Računari u vašoj kući
  • Translated article (translation by me)
• TTL/CMOS logic tester (article by Voja Antonić)
  • Design published in January 1985 issue of Svet kompjutera
• Tone generator (article by Mihajlo Tešević)
  • Design published in Issue 4 of Računari u vašoj kući
• Tone generator 2 (article by Nenad Dunjić, Nenad Balint and Milan Tadić)
  • Design published in April 1986 issue of Svet kompjutera
• Improved graphics (article by Nenad Dunjić and Milan Tadić)
  • Description published in Issue 5 of Računari u vašoj kući
  • Design published in Issue 6 of Računari u vašoj kući
• Galaksija Plus upgrade
  • Galaksija Plus was an improved Galaksija design by Nenad Dunjić, Milan Tadić and Nenad Balint in 1985, it included popular expansion projects like the expanded memory, sound support and improved graphics. More information about the Galaksija Plus is available in the Galaksija Plus section below
  • Starting in January 1986, Svet kompjutera published a series of 4 articles outlining how to upgrade your original Galaksija to the Galaksija Plus model

New expansions

There have been a few modern projects for the Galaksija.

• Galaksija Keyboard Interface – Marko Šolajić’s project, useful for the Galaksija Plus clone and other designs that don’t include a keyboard
• The Galaxino – program loader interface for the Galaksija that can play software recorded in GTP and WAV format. Designed by Michele Perniola from Apulia Retrocomputing

Common problems

• CMOS Z80A signal timings: If you are building a 2020 double layer PCB using a CMOS Z80A you will need to add a capacitor between RFSH and GND pins on the Z80A. There is a spot for a 5nF capacitor right next to the Z80A. If you experience issues with the image having problems with the first row of letters (issue #1 above), or the image being moved around the screen (issue #2 above), try different values for this capacitor, like 3.3nF or even as low as 470pf (0.47nF) until you get a good image. Look here for one example of this issue, and here for another. Read this article for a more thorough background on this issue and an alternative way of dealing with it by building a small circuit

• Image overscan: Depending on your screen, you may have overscan issues where a part of your video output is not showing up on your screen. This is a distinct issue to the Z80A timings adjustment outlined above. To move the image in software, issue: BYTE &2BA8,12 where 12 is the default screen position. Change to 11 to move the image to the left, or 13 moves it to the right. Once you find what works well, you can make that adjustment permanently in ROM B. For a permanent fix, you’ll need to change the value of address 1024 in ROM B to correspond to the number that works well for you (11, 12, 13 etc…). In the original ROM B it is set to “0C” in hex, decimal 12. If BYTE &2BA8,13 worked better for you, then it should be 0D (hex for 13) etc. On my github I have a few versions of ROM B with different values for &2BA8

• Memory issues: Sometimes memory issues will show up as problems in video generation. If you’re using cheap sockets with the RAM chips you may experience issues. I recommend using higher quality sockets, at least for the RAM. You can read about my experience with this issue here

• Issues loading from tape: When building a 2020 double layer PCB version, make sure you solder 3 jumpers for the 3.5mm jack (in between the RCA jack and the 3.5mm jack). If you’re having trouble loading even after you add them, try replacing R14 with 1kohm resistor, and C2 with 1uF capacitor (105 ceramic) – both should be changed at the same time! This is likely caused by using modern transistors which are faster, so changing the resistor + capacitor prolongs the impulse from the tape. Alternatively, you can also try taking out the R14 resistor and leaving the 0.1uF capacitor in place, but this works not as well as replacing both the capacitor and the resistor

Galaksija builds

Here a few examples where folks have documented their process of building their Galaksija.

• Building A Galaksija, Part I: The Tech – the first part of three articles about my experience building a single layer Galaksija
• MCbx Old Computer Collection – single layer Galaksija build log on oldcomputer.info. A neat build including a homemade PCB
• Building the Galaksija on YouTube – Hristijan L recorded himself building a 2020 double layer Galaksija in a series of videos
• Building the Galaksija Plus Computer – Retro Bertie’s video about building a Galaksija Plus clone
• Making a Galaksija – Galaksija plus clone build log on kernelcrash.com

Galaksija Plus

There are many references to the Galaksija Plus and its clones online and in this FAQ document, but the history of the Galaksija Plus is not well documented. “Galaksija Plus” refers to both hardware improvement projects (as outlined above in the Expansions section), as well as a commercial model that was offered for sale.

Galaksija Plus was an improved Galaksija design by Nenad Dunjić, Milan Tadić and Nenad Balint in 1985, it included popular expansion projects like the expanded memory, sound support and improved graphics. Galaksija Plus was described in issue 8 of Računari u vašoj kući magazine (October 1985), October 1985 issue of Svet kompjutera and November 1985 issue of Svet kompjutera.

Starting in January 1986, Svet kompjutera published a series of 4 articles outlining how to upgrade your original Galaksija to the Galaksija Plus model.

There are multiple Galaksija Plus clones (as outlined above) and emulators like MAME support it as well (use this zip file).

Additional resources

• Hacking the Digital and Social System – Voja Antonić’s article on Hackaday about the origins of the Galaksija computer
• How to use more than 100% of program memory – Voja Antonić’s updated 2020 double layer Galaksija PCB design
• Building A Galaksija, Part I: The Tech – the first part of three articles about my experience building a Galaksija
• Building A Galaksija, Part II: The Origin Story – the second part of three articles about my experience building a Galaksija
• Building A Galaksija, Part III: Audio Cassettes, Magazine Listings And Radio Waves – the third part of three articles about my experience building a Galaksija
• Revolution During Disintegration: Lessons from a Short History of Yugoslav Computing (HOPE 2022) – a recording of the talk I gave at A New HOPE conference on 7/24/2022 in NYC.
• My Galaksija github repo – features most of the files referenced in this post
• Galaksija programs on retrospec.sgn.net
• Tomaž Šolc’s series of posts about the Galaksija – Tomaž Šolc documented many in depth technical aspects of the galaksija, including the video generator circuit, character generation, and more (in English)
• Tomaž Šolc’s Master’s Thesis on building a CMOS Galaksija replica (in Slovene)
• Tomaž Šolc’s Master Thesis on building a CMOS Galaksija replica (English translation work in progress)
• 29C3: The ultimate Galaksija talk – Tomaž Šolc’s talk about the Galaksija at Chaos Communication Congress in 2012
• Galaksija talk at KiberPipa hackerspace – Tomaž Šolc’s talka bout the Galaksija (in Slovene)
• Tomaž Šolc’s Galaksija development tools
• Galaksija ROM A dissasembly by Tomaž Šolc (in English)
• A Galaksija Not So Far Away: The Story and Legacy of the Galaksija Computer – an episode of the Remembering Yugoslavia podcast featuring Voja Antonić, Dejan Ristanović and myself
• Galaksija documentary film (in Serbian with English subtitles)
• Galaksija Replica Project – description of the Apulia Retrocomputing Galaksija replica project
• Apulia Retrocomputing documentation – documentation files for the Apulia Retrocomputing Galaksija replica project
• Il ritorno del #Galaksija! [Galaksija is BACK!] – Youtube video about the Galaksija replica project at Apulia retrocomputing (in Italian with English subtitles)
• The Ages of Galaksija – a short documentary film by Guido Cauli in partnership with Apulia Retrocomputing (in Italian with English subtitles)
• специалист-пк.рф/Galaksija – the “Russian website about the Galaksija” with gerber files, schematic and ROM files. Originally at spetsialist-mx.ru
• Hindicki Ferenc’s Galaksija offline website – This is a large zip file with a huge amount of Galaksija materials that can be browsed as an offline webpage (start by opening index.html) as well as the Jsgalmin emulator (mentioned in the Emulators section). A live version is available at http://galaksija.epizy.com
• Računar Galaksija – moja priča, Dejan Ristanović – oral history of Galaksija, by Dejan Ristanović (in Serbian)
• Oral History of Voja Antonić – interview with Voja Antonić at the Computer History Museum in Mountain View, CA
• Miodrag Milanović’s galtape utility – the best tool to convert WAV files to gtp files

The post Making a Galaksija computer FAQ appeared first on Vlado Vince.

It’s MARCHintosh! The month of March when the vintage computing community shares projects, creations and other content related to the Macintosh, in the vein of DOScember and SEPTandy. It’s almost halfway through March 2023, and due to my increasingly busy work schedule, I don’t believe I’ll get to do a project of my own. However, back in 2022 I did do a small #MARCHintosh project on Twitter, documenting the brief history of the Mac in Yugoslavia, primarily through the lens of its appearance in the (then still nascent) Yugoslav computing magazines. Enjoy!

This mini project started when I found out that the first Macintosh commercials (other than the Super Bowl one) were first aired during 1984 Sarajevo Olympics coverage. I don’t have any actual examples of this, though it shouldn’t be too difficult to find a recording of ABC’s 1984 Olympics coverage. The first lead for this came from Steve Jobs’s Apple Macintosh premiere with the Mac team, at the Boston Computer Society General Meeting, on January 30, 1984. Steve Jobs introduces the Macintosh, followed by a panel including Steve Capps, Andy Hertzfeld, Randy Wigginton, Bill Atkinson, Bruce Horn, Burrell Smith, Owen Densmore, and Rony Sebok. Steve Jobs doesn’t explicitly call out Sarajevo, but after showing one of the commercials, he says:

“So if you watch the winter Olympics, you’ll be seeing those about two or three times a night”

Unsurprisingly it took quite a long time for the Macintosh to make its way to Yugoslavia, aside from TV commercials during the Sarajevo Olympics. When the Mac was released in Jan 1984, Yugoslavs were still unable to directly purchase even the cheapest Western micros like the ZX Spectrum! The same month the Macintosh was released, Yugoslavs went to the newsstands to buy Računari u vašoj kući magazine, some of them learning about computers for the first time. In there they found instructions on how they can build a computer themselves: the Galaksija.

In the July issue of Računari u vašoj kući (only the second issue of this magazine!) there is a brief mention of the Macintosh in a short report from COMDEX 1984, humorously titled “nothing new in the West”:

There’s also mention of the Mac in the inaugural issue of Svet kompjutera in October 1984. This is 10 months after the Mac came out! News travels slow at this time. You can see the disconnect further down in the article. While the author (rather correctly) predicts that the Mac will be at least as influential as the Apple II was, they speculate the 256K version will be released in 1985, even though the 512K Mac was actually released in September:

If we pay attention to ads we’ll start to notice Velebit Informatika, which will soon grow into the main Yugoslav distributor for Apple:

Even though they sold different domestic and foreign computers, Velebit informatika increasingly focused on Apple in particular:

We can follow the actual arrival of the Mac to Yugoslavia through their ads. Here’s an event in May 1986 where Velebit Informatika plans to present the Macintosh Plus:

We can also see that they offer 30% discount for education and 20% discount for “working organizations”, the type of small and medium size organizations under Yugoslav self-management system:

But Velebit was more than just a reseller — here they are marketing an Apple II configured with a domestically developed electronic health record system called Hipokrat:

From what I can tell, the first review of the Macintosh in the Yugoslav press was in Moj Mikro in June 1984. The reviewer, Žiga Turk (who was later a minister in Slovenian government!), got to try it at IFABO expo in Vienna. He also mocked up Ljubljana’s Radio Student logo in MacPaint.

This same review was re-published in January 1985 when Moj Mikro started publishing a Serbocroatian edition. Žiga was impressed with the Mac, but somberly noted at the end that with current restrictive import policy it will be very difficult for anyone in YU to get one.

Another interesting ad – it wasn’t unusual for computer stores in Italy or Austria to advertise in Yugoslav magazines:

Another YU Macintosh memento. A “mini poster” from Feb 1985 Računari u vašoj kući. It might have been unobtainable for most, but you could still put it on your wall:

Velebit informatika published localized Mac applications through the 80s:

And on January 20th 1987 they opened their first Apple showroom in Belgrade, Yugoslavia’s first “Apple store”:

The Mac never achieved widespread usage in Yugoslavia, but these examples show that very soon after its announcement in 1984 it started showing up in the media. While there were very few Macintosh home users, Velebit informatika’s software ads show that there was an emerging market for the Mac in the business environment.

The post Macintosh in Yugoslavia – a belated #MARCHintosh2022 retrospective appeared first on Vlado Vince.

That was more or less the end of my Pocket PC experience. In the late 2000s I did use an HTC P4350 “smartphone” running Windows Mobile, which was basically Windows CE from the PDA days, but it was just a phone, a brief interlude before everything became iPhone or Android.

I held onto the broken iPaq for many years, and then got rid of it during one of my cursed purges of old devices. A horrible mistake that has been corrected a few weeks ago when I got the same iPaq model on eBay for $10 + free shipping. The iPaq arrived fully functional, sans the battery that would no longer hold charge, and the stylus release button that would no longer hold the stylus in place. But otherwise it worked!

The iPaq came with the original Windows CE 3.0 (also known as Pocket PC 2000) OS. I enjoy the Windows 98 era display style of this version, but back when I used this iPaq I upgraded it to Pocket PC 2002, so naturally I decided to do the same with this one. I realized that playing with this device brought similar type of tech nostalgia that folks who grew up in the 80s have towards Commodore 64s and Apple IIs — while I enjoy using those, they are not actually a part of my own past. The iPaq, however, most definitely is.

So, I went ahead and obtained the Pocket PC 2002 upgrade iso from the Internet Archive. Back in the 2000s you were supposed to obtain a license from Compaq to perform this upgrade. Regardless, I was able to easily obtain it, even on a pre-broadband ISDN 64kbps dial up connection. And this is where this post moves from personal history into troubleshooting documentation. If you are interested in upgrading 20 year old PDA’s OS in 2023 and beyond – don’t be lazy and attempt this using a Windows XP virtual machine. Go ahead and perform this upgrade using real hardware.

I was lazy.

The install went fine and detected my iPaq through the Virtualbox USB passthrough, and it began the ROM flashing process. It made a backup of the current ROM (this will come in handy later on!) and started to flash the new OS. At this point however, the iPaq rebooted into the ROM flashing mode, and Virtualbox lost the connection. The ROM flashed failed, while the iPaq remained stuck on the ROM flashing screen. I naively unplugged it from its craddle, which caused it to lose power (with the dead battery and all), and at the next attempt it failed to boot. The screen would just turn black, with no backlight, with the battery light flashing to show that it had power.

Needless to say I was disappointed, but also sort of excited. I laughed at myself, realizing that 20 years later I made the type of mistake that my 12 year old self managed to avoid. Additionally, fixing these types of issues on old tech is what brings me joy anyway, so I embarked on a hour long journey to figure out how to restore this device that is so dear to me. I’ve found that troubleshooting hardware and software from the late 90s – early 2000s tends to be much more difficult than dealing with the stuff from the 80s. It’s recent enough that it tends to be quite technically complex, but there’s still limited attention from the vintage computing community, so you mainly depend on contemporary accounts from the time period when folks actually used this stuff. And forum posts and blogs from the early 2000s are increasingly disappearing from the web. While the Internet Archive helps a lot, ultimately you do tend to depend on Google to find leads in the first place, and Google has been increasingly unreliable as of late.

However, within the hour I more or less figured out how to get out of this. Below is step by step:

• Make sure you have a copy of your iPaq’s ROM in .nbf format. Mine was IPAQROM177.nbf
• Get the host11.exe flashing utility (this was the crucial ingredient)
• remove the first 32 bytes from the backed up ROM file in .nbf format and rename it to nk.nb0. Here’s the command if you want to use dd: dd bs=32 skip=1 if=IPAQROM177.nbf of=nk.nb0
• Place the host11.exe and nk.nb0 on the computer you’ll use to restore the iPaq

• Reset your iPaq by holding down the circle (action pad) and the power button and hitting the reset button. You should then see an image of a parrot with the title “Remote USB Update”. It took a few tries to get this to show up
• Connect the iPaq via USB
• Start host11.exe

• The flashing will take about 2-3 minutes and your iPaq will boot with the original ROM

I got these instructions from a long ago lost page from handhelds.org. The most difficult thing to track down was host11.exe, which is why I’ve uploaded it here in case anyone needs it.

The reflashing was successful, and I quickly got back to the original Windows CE 3.0 screen. I hope this is helpful for anyone playing around with this device in the 2020s. I’m definitely enjoying this pocket PC from my personal history!

The post Reliving my past and de-bricking a Compaq iPaq 3650 in 2023 appeared first on Vlado Vince.

Edited May 16 2023

Orao was a Yugoslav 8-bit computer developed by Miroslav Kocijan for PEL Varaždin in 1984. Orao was common in elementary schools, mostly in Croatia, through the late 80s. Since 2021, I’ve been learning more about this microcomputer and developing hardware and software for it. While most of the information here is available on my Orao github repo, this page contains all the relevant information related to the Orao WiFi RS232 adapter, which I developed in 2022.

The Orao has a built in RS232 TX meant for use with a printer, but no built in RX. In order to enable two way serial communication I built a 6551 based RS232 adapter, which was later modified to include a Raspberry Pi Zero W, which serves as an emulated modem, using tcpser. I also wrote a simple terminal emulator that needs to be loaded from tape, called Orao Terminal. Below is a quick list of resources for use with the Orao WiFi RS232:

Hardware:

• PCB Gerbers file
• Parts BOM
• Raspberry Pi SD card image with tcpser installed. Uncompress with unxz and copy to SD card with dd. To connect to WiFi, create a file “/boot/wpa_supplicant.conf” and fill out with your wifi SSID and password:

country=US

ctrl_interface=DIR=/var/run/wpa_supplicant GROUP=netdev

update_config=1

network={

ssid="MY_SSID"

psk="MY_PASSWORD"

}

Software

• Orao Terminal .wav program
• Orao Terminal assembler

Instructions

Hardware

Orao WiFi RS232 is built around the 6551 asynchronous communication adapter (ACIA). The ACIA is placed at memory addresses A000 (data register), A800 (status register), B000 (command register) and B800 (control register). The addresses are decoded with the 74HC10 triple NAND gate, and the RS232 voltage is adjusted through the MAX232 IC. 1.8432Mhz oscillator is used to set the serial bitrate. A Raspberry Pi Zero W can be used to emulate a modem via the tcpser program, but a DB-9 port and an open header with exposed RS232 signals can be used as well. An additional 74HC00 IC is used for the status LEDs and to amplify the PHI2 signal.

Since the Orao doesn’t expose 5V on the expansion connector, an external power adapter is needed. You can either use the barrel jack or a micro USB port to provide 5V DC.

Orao WiFi RS232 has three status LEDs:
• Blue (ON) – lights up when external power is connected
• Red (RXD) – blinks when data is received
• Green (TXD) – blinks when data is sent

Orao WiFi RS232 uses a Raspberry Pi Zero W to emulate a modem, using the tcpser program. In order to communicate with the Orao via the 6551 ACIA, RPi’s UART is directly connected to the 6551 (since it uses TTL levels). While the Pi Zero W works the best because of the size, you can also use any regular size Raspberry Pi newer than the Pi 2. Before using a different Pi, check that the pinout is the same as on the Pi Zero.

Orao WiFi RS232 assembly instructions are available here (in BCMS).

Schematic

Software

Orao Terminal supports serial communications over 300, 2400 and 4800 bps. Orao Terminal supports ASCII encoding (instead of the native YUSCII encoding normally used on the Orao).

The program is fully written in 6502 assembly. It’s loaded from BASIC using the LMEM “” command.

Afterwards, you can choose the mode:

1. WiFi Modem
2. RS232

Selecting WiFi modem will set the speed to 4800 (default used by the Raspberry Pi Zero W in modem emulation mode). Selecting RS232 will show another menu with additional speeds.

Notes:

Since the Shift key functionality is not implemented yet, the 4 PF buttons have the following uses:

Shift key has now been implemented and works as it should with capital letters and symbols. Functions for the PF buttons are below:

PF1 = clear screen

PF2 = not currently implemented

PF3 = not currently implemented

PF4 =not currently implemented

The program uses hardware CTS/RTS flow control. Flow control is necessary because of the limited screen space. Flow control will automatically stop RX once all 32 rows are filled. Pressing the CR button will clear the screen and continue receiving data.

At 4800bps, some responses from tcpser might be too fast (e.g. responding OK to AT command). This only applies to responses immediately following pressing CR. Lowering speed to 2400 or 300 may resolve this issue.

Edit: issues with fast responses have been resolved in Orao Terminal versions 0.32 and higher.

Raspberry Pi Zero W prep

You can download the Raspberry Pi SD card image with tcpser installed. Uncompress with unxz and copy to SD card with dd. To connect to WiFi, create a file “/boot/wpa_supplicant.conf” and fill out with your wifi SSID and password.

On a fresh installation, the following steps need to be performed on the Raspberry Pi:

• Download the Raspberry Pi OS Lite from raspberrypi.org
• Raspberry Pi OS images now come compresed as .img.xz. You can use unxz to decompress the .img
• Insert the SD card and copy the image using dd
• When the image is copied, mount the /boot partition on your computer
• First we need to create a new account. As of 2022 the default pi/raspberry account is no longer active on the default Raspberry Pi OS image
• Generate an encrypted version of your password using openssl (here “orao2022” is the desired passowrd

echo 'orao2022' | openssl passwd -6 -stdin

• Create a file “/boot/userconf” and enter your credentials in the format “username:encrypted- password”, e.g.:

pi:$6$mc1/HL5CSKUJcfgL$.6Il3lJOs9uKRfFJqvmJwaHU8n3A3OOKqj.abb.eAf0/nR.B8urz4KJnDDMIgaciBvGhEon7nH1HRZC.2EKLD1

• This will create a user “pi” with password “orao2022”
• Now we need to enable SSH. Create a file “/boot/ssh”
• To connect to WiFi, create a file “/boot/wpa_supplicant.conf” and fill out with your wifi SSID and password:

country=US
ctrl_interface=DIR=/var/run/wpa_supplicant GROUP=netdev
update_config=1

network={
	ssid="MY_SSID"
	psk="MY_PASSWORD"
}

• Lastly we should enable the built in UART and disable Bluetooth (because RPi uses the UART for BT communication by default). Edit /boot/config.txt and add the following two lines:

enable_uart=1
dtoverlay=disable-bt

• We can now put the SD card into the Pi and boot it up
• The first boot might take a bit longer
• Once the Pi is powered on, you can check your router for its IP, or use nmap. Once you have the IP, SSH into the Pi using the pi/orao2022 credentials (or whatever you used)
• We should first run raspi-config and set the locale so we don’t get any errors
• Run sudo apt-get update, and then sudo apt-get upgrade so we have all the updates
• Once the updates are installed, we need to install tcpser, rpirtscts and telnetd
• First install git with sudo apt-get install git
• Then install tcpser:

git clone https://github.com/FozzTexx/tcpser.git
cd tcpser/
make
sudo cp tcpser /usr/bin

• Now let’s install rpirtscts. This utility is needed to enable hardware flow control on the built in UART

wget http://ftp.podsix.org/pub/pimodem/mholling-rpirtscts.zip
unzip mholling-rpirtscts.zip
cd rpirtscts-master
sudo cp -p rpirtscts /usr/local/bin

• Lastly, let’s install telnetd. Running a telnet server on the Raspberry Pi will let us control the Pi directly from the Orao by using the ATDT127.0.0.1 command. Use telnet on a networked device at your own risk.

sudo apt-get install telnetd

• We should now add the following two lines to /etc/rc.local to automatically start tcpser and rpirtscts on boot. The lines should be added after fi, but before exit 0

/usr/local/bin/rpirtscts on
tcpser -d /dev/ttyAMA0 -s 4800 -tsS 

• We can now shutdown the Raspberry Pi and install it on the Orao WiFi RS232. If we ran all the commands correctly, it should start in modem emulation mode and we should be able to use it to connect to networked resources using ATDT and other common modem commands

Acknowledgments

This project is a result of many hours of work, but also at least a dozen other projects that served as inspiration or direct resources that helped me make the Orao WiFi RS232. They are listed here in no particular order:

• Josip Perušanec’s Orao 2007 clone – before I obtained a real Orao computer, I put together a fully functional Orao clone designed by Josip Perušanec in 2007. His project is available here. I built my first Orao hardware project back in 2021 in order to use a PS2 keyboard with the Orao 2007 clone. Documentation of that project is available on my Orao github repo
• The late Lee Davison’s 6551 circuit, hosted by Hans Otten, was my initial inspiration on how to build this adapter
• I also referenced Dirk Grappendorf’s blog post about implementing 6551 in his home computer
• Another resource I referenced is Michael Billington’s blog post about the 6551. This post was especially useful for me to understand how to implement address decoding
• Steve Mansfield-Devine’s blog post about implementing the 6551 for his Zolatron computer was very useful too, especially to understand the differences between different 6551 chips
• Manuals and books about the Orao computer were indispensable, especially the two books published online by RetroInfo fairly recently. These resources are only available in Croatian. All the materials I used are hosted on my github.
• Željko Perić’s Orao Basket. Željko’s set of Orao development utilities was very helpful in writing the Orao Terminal emulator, especially the assembler program and the bitmap image converter
• tcpser – this handy modem emulator utility is what makes the Orao WiFi RS232 useful. I used this particular fork of the program originally written by Jim Brain
• Vince Weaver’s explanation of the built in UART on the Raspberry Pi was very helpful
• Matthew Hollingworth’s rpirtsrtc enabled me to use flow control on a Raspberry Pi
• Big thanks to Juraj Mucko who kindly gifted me two fully functional Orao computers
• Shout out to PEEK&POKE museum in Rijeka where I got to see the Orao for the first time. Visiting PEEK&POKE is what got me interested in vintage Yugoslav computers in the first place
• And lastly a thank you to Miroslav Kocijan who designed the Orao back in 1984!

The post Orao WiFi/RS232 and Orao Terminal emulator appeared first on Vlado Vince.

As I write this (in mid-November), it’s hard to believe that it’s only been 3 weeks since the surreally stupid moment when Elon Musk triumphantly walked into Twitter offices carrying a sink (I guess with the intention of making thousands of 12-year-olds laugh with by performing the “let that sink in” meme in real life). In those three weeks, the currently richest man in the world fired the board of directors, most of the executive team, over five thousand Twitter employees (50% of the workforce) and an unknown amount of subcontractors who performed extremely important work of moderating the social network. Since walking into Twitter offices, Elon has made a bunch of seemingly frantic decisions on a daily basis, from demanding that programmers print out all the code they’ve written in the past months (so that, I guess, the loyal engineers he brought in from Tesla and Space X can evaluate it based on length), to the spectacularly chaotic introduction of paid verification, which led to multimillion-dollar losses for companies that suddenly found themselves in a situation where anyone with eight dollars could open a Twitter profile that looks as official as theirs.

But all this is the culmination of a months-long struggle for Elon to get out of buying the world’s most influential social network. It’s unclear exactly why he decided to pay $44 billion for Twitter last spring (and it’s $44 billion he didn’t have), but shortly after his original announcement, he began looking for reasons to cancel the sale. On the other hand, Twitter’s supervisory board relatively quickly accepted Musk’s offer to completely take over the company, which until then was publicly traded. This is not surprising, given that Twitter is known for losing money – in 2019 it ended the year with a profit for the first time ever, so from the perspective of the board, this offer is probably the only one that could bring a return to the investors. Musk then changed his mind, and spent months trying to get out of the sale, including his attempt to convince a court that he had been duped and that Twitter had not given him accurate information about the presence of “bots” on the platform. That lawsuit failed, and that’s how we got to him entering the Twitter building in San Francisco at the end of October. Musk eventually raised the money with the help of a loan backed by Tesla shares, as well as a number of partners, including the Saudi Prince Alwaleed bin Talal and the Qatari sovereign wealth fund.

After mass layoffs and a completely chaotic first couple of weeks, it’s still completely unclear what Elon’s actual goal is. He announced that he wanted to turn Twitter into an “oasis of free speech”, and the first results of his moves show that this freedom is reserved only for his followers, who now mainly consist of cryptocurrency traders (crypto is another one of Elon’s “hobbies”) and right-wing trolls (like the user @catturd2 with whom Elon corresponds and discusses the moves he plans to make in his new company). But although the removal of verified users who ridiculed him shows that Elon’s understanding of free speech is more similar to that of the Gulf dictatorships that helped finance his purchase of Twitter than, for example, to the definition found in the US Constitution, it is not entirely clear whether Elon has any constructive intentions for Twitter or whether he is straight up trying to destroy a social network that has had an unprecedented global influence over its fifteen years of operations. Regardless of whether Twitter ceases to function by the end of the year, which is absolutely possible at the moment, or lives on as an oasis for conspiracy theorists and Tesla fanboys for years to come, the damage Elon has already done is probably beyond repair. Hundreds of thousands of users, if not millions, have already left the network, in an exodus the likes of which we haven’t seen on the Internet in years. It is completely impossible to speculate what economic, social, geopolitical and other consequences this process will have, especially in the context of actual freedom of speech, the reporting on war crimes and human rights violations. For years, Twitter has been the main channel for reporting from conflict zones, from the Arab Spring to the war in Ukraine.

But this text is not about that, and it is not about Elon Musk either. Books and documentaries will probably be written about the background and results of this moment (personally, I hope someone makes a documentary along the lines of the one about the Fyre Festival). Instead, this text is very personal. I am writing it because behind all the current news stories, behind everything bad and traumatic, and we know that bad things on Twitter did not start with Elon Musk, there are countless stories of us who have been on that social network for years – we’ve socialized, met up and built our communities over there. We are not unique in this respect – online communities have been growing, living and falling apart since the first BBSs (Bulletin Board System) from the early 1980s. But before we look back to learn something from those earlier attempts, I’ll tell you briefly what Twitter has meant to me, and how it’s connected me to new and interesting people over the years.

In 2006, when Twitter was launched, I was in high school. By this point I’ve been active in various internet communities for several years, first in local forums about science fiction (whoever has been there will know which ones), then on blogs, which exploded in Croatia around that time. When I originally created a Twitter account, in September 2007, no one really know what kind of social network it would be or how to communicate on it. They called it a “microblogging” platform, so we all imitated our blogs from that time, in 140 characters (the original limit of one tweet, based on the length of a single SMS message). But even then you could see what, from my perspective, remains the key element for Twitter’s succes. Unlike forums and blogs, on Twitter we were all talking to everyone else. Not literally, of course, but the network expanded organically through the follow system so that we would meet new people through the people we already follow, regardless of particular interests or smaller groups. When I was on forums I talked to Star Trek fans. On blogs, I hung out with photographers because I had a photo blog. When I got on Twitter, suddenly I was hanging out with everyone! In the beginning it was easy because there were very few of us, especially on Croatian Twitter. We connected spontaneously, hung out at tweet-up meetings, and I still keep in touch with many of them, even though they left Twitter a long time ago.

In 2010, I moved to America. For the first few years, I didn’t use Twitter that actively anymore. However, it was my connection with people in the Balkans, an opportunity to communicate regularly in my own language, but also to follow what was happening in the country I left. After moving to New York in 2014, Twitter was crucial in getting involved in cycling and urban activism in that city, but even though my profile became bilingual like my “real” life, my “native” Twitter community remains the most important one to me. Of course, there is not just one “domestic” Twitter – even in the small area of former Yugoslavia there are at least one hundred thousand active users, probably more. But the opportunity to continuously meet new people, spontaneously, in my own language, was extremely important to me. One element of emigration, which in my experience is not talked about much, is the stagnation of new friendships after a long absence from the country you left. Of course, if you come home relatively regularly (in my case it’s once or twice a year for a couple of weeks), you will meet some new people. However, opportunities to develop new relationships are very limited. Twitter changed that dynamic for me – many of the people I met in recent years in Croatia were initially “mutuals” (people who follow each other on Twitter). I’m not mentioning this as some kind of unique experience – in 2022, people will meet each other on the internet through all different networks and apps, but I’m pointing out that for a specific group of people, Twitter was a place that despite its toxicity continuously encouraged positive relationships that developed into friendships .

Some of us started preparing for the move back in April, when Elon first announced the purchase. We had no way of knowing exactly what was going to happen, but based on his reputation, we assumed he wasn’t planning anything good for our network. Many opened profiles on Mastodon, a decentralized platform similar in functionality to Twitter, but ideologically and infrastructurally completely different from it, more similar to decentralized communities from the earlier stages of online communications. For some it was the first time they were going through a collapse of a network, others have been through multiple such events, from MySpace, Tumblr (which is potentially experiencing a second renaissance through the collapse of Twitter), MSN, to countless forums, Usenet groups and BBSs. There wasn’t a consensus on whether to move elsewhere, and that’s exactly the reason why I don’t believe that the communities created on Twitter will survive in the form they functioned until October 2022. Some people will stay on Twitter until the end. Some will go to Mastodon, some to Tumblr, some younger and more creative ones might go to Tiktok, some to private Signal and Telegram groups, some to Discord. Some simply won’t go anywhere, maybe they will devote themselves to friends in real life. Some will leave and then come back. Like any other disintegration, this one very much feels like balkanization, to use a problematic but seemingly appropriate term. In an academic context, balkanization is used to describe the fragmentation of territory into smaller parts. And after everyone goes their separate ways, it will take a long time to build relationships again.

Unlike many geopolitical meltdowns, this one may have some positive consequences. In the mid-2000s, the Internet was going through its Web 2.0 phase. The emphasis was on platforms that allowed users to actively create and share their own content – YouTube, Flickr (once the main photo-sharing platform), blogs, Facebook, Twitter, they all were created during the crucial period when the Internet we know today was born. At the beginning, the emphasis was on the openness of platforms, but also the openness of technology, through open APIs (interfaces for communication between platforms) and technologies such as RSS (which we used to follow blogs). However, this gave way to closed centralized platforms that later on grew into companies worth tens of billions of dollars. When talk of decentralization began again in the 2010s, cryptocurrencies and blockchain took over. Instead of decentralization in a social and political sense, the ideology that developed in crypto communities saw social relations exclusively in the form of monetary transactions. Coincidentally (or not?), along with the collapse of Twitter, we are currently witnessing the implosion of FTX, one of the largest cryptocurrency exchanges. Given the far-reaching consequences of this and other crashes in the cryptocurrency markets, we see that their promises of decentralization were false – the crypto movement centralized financial power in companies controlled by a small group of people, just like the financial system it wanted to replace.

But maybe something is changing. Among a bunch of options, I decided to find and build my community again on Mastodon. Although we have already described it as a decentralized Twitter, Mastodon is relatively difficult to explain without getting into technical details and jargon that many people do not understand. Therefore, I will try to explain it in social and economic terms. Mastodon is only one part of the fediverse environment based on open standards. It’s a network that, in terms of infrastructure and society, rests on the will and effort of users to manage it, as opposed to networks run by large companies which give us a “free” service in exchange for our data and its manipulation through secret algorithms. In practical terms, this means that Mastodon is similar to the early internet and pre-internet, when online communities like BBSs were created through the work of a few people who wanted to communicate with each other. Some of them would maintain the technical infrastructure, servers and phone connections, while others would moderate the communication to keep the community healthy and functional. The problems that Twitter is dealing with today, or rather, problems it was dealing with until Elon fired the moderators – harassment, homophobia and transphobia, racism and an increasing number of Nazis – have been present since the first online communities. But operators of those communities learned even then that unconditional freedom of speech does not exist – if toxic and hateful members are tolerated in a community, they will destroy the community and drive out everyone else. In the period of dominance of centralized social networks that has lasted for almost twenty years now, companies like Facebook and Twitter have tried to deal with this problem through a combination of algorithmic and manual moderation, by constant ideological balancing to (unsuccessfully) avoid accusations of bias. The results are disappointing: we’ve known for years that algorithms tolerate racism and regularly discriminate against people who fight against it. Corporate social networks have been implicated in key global scandals of our time, like the documented Russian interference in the 2016 US election.

Mastodon and the fediverse promise us something completely opposite. Launched in 2018, Mastodon has grown slowly. In the world of technology and economy dominated by the ideology of the Silicon Valley where the choices are either to grow or to disappear, such an approach is completely anachronistic. Instead of having acres of centralized data centers in the desert like Facebook and Twitter, Mastodon is more like the networks of the 1980s and 1990s where communities maintained and moderated their own servers. But what makes it modern and ready to replace Twitter is the protocol that allows all these communities to communicate with each other. It’s built on the principle of subsidiarity – the responsibility for community building – in the technological-infrastructural and social sense – is transferred from the level of a global corporation to user groups. Well-managed communities then cooperate with each other to keep the entire network healthy and functional. If a group of Nazis appear on one server, a good moderator will kick them out. If that moderator fails and allows their community to turn into a community of Nazis, the other server’s admins will block them. At least that’s how things should work. Whether Mastodon is ready for a major surge of users after the Twitter implosion remains to be seen. There are several challenges, technological and social, but also practical – Mastodon is still relatively complicated for new users. The server concept is abstract to folks after 20 years of centralized services, and many users will find the techno-utopian atmosphere off-putting at first. Fortunately, there are already many people on Mastodon who understand these challenges and are actively trying to welcome new arrivals. Although the majority of “domestic” users are scattered across various servers, there are several regional ones – toot.si, mastodon.hr and balkan.fedive.rs. Those of us who recently arrived from Twitter welcome new users via the hashtag #bivšitviteraši (former Twitter users) and offer help to people who are trying to adapt to the new platform. In general, there is a very optimistic and empathetic atmosphere on Mastodon right now, so unusual on today’s internet. A good example is the attention paid to accessibility, for example, by adding textual descriptions of images for people with low vision.

No one knows what’s next for Twitter. Every day there are new crazy news – it is possible that by the time this text is published, this analysis will be completely out of date. We live in an extremely dynamic and unpredictable period. The pandemic is still going on, there is a war in Ukraine, climate change is getting worse, and it seems that a big economic crisis is coming. It’s hard to predict how Twitter’s changes will affect all these political crises, but I believe that Musk’s purchase of Twitter will become a significant historical moment. But maybe along with the story about major global events, we can start writing a smaller one, about our online communities that we started to rebuild together and independently from the crazed billionaire who carries a sink. As for me, I’ve been @mejs on Twitter for the past 15 years. Now I’m @mejs@mastodon.vladovince.com on Mastodon. See you on the internet!

The post Ode to Twitter appeared first on Vlado Vince.

I desperately longed for some type of collective memory and grieving, not just for him but for all the millions lost to the virus. And there was nothing. In Croatia and in America, we collectively shrugged and chose to look the other way and take off the masks.

I posted photos and memories of my grandfather for 100 days after he died. It was therapeutic, but what really helped was support and mutual care from the many people who responded. This was important to me. It helped me keep going. I wouldn’t have had that without Twitter. What follows is the verbatim copy of those posts.

In memory of my grandfather Josip Gobac (1932 – 2021) I’ll be posting one photo of him every day. He not only led an amazingly unique life, but he also left a ton of wonderful photos. I’ll start with a fairly recent photo of him filming the Degenija flower on Velebit in 2012

Day two of memories of my grandfather Josip Gobac (1932 – 2021). This one is from his art school days at Škola primjenjene umjetnosti. Hanging out on the dorm roof in 1955. My grandpa is on the right, standing.

Day 3 celebrating the life of my grandfather Josip Gobac (1932 – 2021). Taken in front of Stazione Centrale in Milan. He was there in 1954 attending X Triennale di Milano. The note on the back says the woman is Viviana Bernardi, but I’m not sure who she is (en.wikipedia.org/wiki/Milan_Tri…)

Day 4 of memories of my grandfather Josip Gobac (1932 – 2021). Switching gears a bit now and taking us back to 2012. Reading the newspaper in Ashland, Oregon.

Day 5 of memories of my grandfather Josip Gobac (1932 – 2021). Location and year unknown, but it’s likely close to Kutina and taken in the late 60s. Slide photo.

Day 6 of memories of my grandfather Josip Gobac (1932 – 2021). I don’t know where this photo was taken or when, but I assume it’s the late 50s or 60s. It’s also a somewhat rare photo of my grandpa at (what I assume) is the seaside. He never got fully comfortable with the sea.

Day 7 remembering my grandfather Josip Gobac (1932 – 2021). Let’s go to the 80s — here he is at the Battery in NYC in 1986 with Marinko Čubrilo and Dubravko Šarić. He didn’t really like New York very much and described it as noisy and messy (still correct).

Day 8 remembering my grandfather Josip Gobac (1932 – 2021). Continuing with the same US trip in 1986, here he’s boarding a streetcar in New Orleans. Unlike NYC he really liked that city.

Day 9 remembering my grandfather. We’re staying in the US but fast forwarding 25 years to 2012. My grandpa probably ate fewer than 5 hamburgers in his life, and this is one of those times. Somewhere in Northern California, I think. He was a good sport about it.

Day 10 remembering my grandfather Josip Gobac (1932 – 2021). This is how I imagine him when I think about him — with his camera in the mountains. This is from a hike we did in 2007 on Velebit, on our way to Crnopac.

Day 11 remembering my grandfather Josip Gobac (1932 – 2021). This is from the “zu Gast bei” show on Hessen Drei TV channel in Frankfurt (now called hr-fernsehen). The note says 1980, but it might be from a few years later. Grandpa is right in the middle.

Day 12 remembering my grandfather. Here’s a photo of the two of us on a bridge over river Lika, near Kaluđerovac. Taken in June 1997.

Day 13 remembering my grandfather. A double feature today with two photos from a climb to the top of Triglav, the highest mountain in Slovenia. I was supposed to join him, but for some reason I didn’t make it on this trip. I hope to climb it for him one day.

Day 14 remembering my grandfather Josip Gobac (1932 – 2021). This is his class photo taken at his 8th grade graduation in 1946 (“Mala matura”). His schooling was interrupted during the Ustaše/German occupation of Kutina. Grandpa is in the top row, second from the left.

Day 15 remembering my grandfather. This is my grandpa’s wedding to my grandmother Slobodanka in 1962. Quite a formal affair! I love my grandma’s dress, and how my grandpa’s suit looks a bit awkward (later on he always had such a well-cut suit!). He’s basically my age in this one

Day 16 remembering my grandfather. 1969 in front of the Most slobode bridge in Zagreb. Presumably next to the former cable ferry that operated there before the bridge was opened in 1959?

Day 17 remembering my grandpa. 1995 at Croatian Television (HRT) main building. He managed the construction of its building complex while it was still called Televizija Zagreb. Like many journalists and workers he was forced into early retirement in 1991.

Day 18 remembering my grandpa. In 2016 he received a special award by the town of Popovača for his contributions. Since 2012, he produced yearly short films about the new developments in town and he gifted the films and the rights to the town for free. I love how happy he is.

Day 19 remembering my grandfather. January 1st 1980 in Nefta, Tunisia. My grandmother Slobodanka loved new year celebrations so they often traveled. Grandpa is in the back, grandma is next to him and my mom is in the red sweater.

Day 20 remembering my grandfather Josip Gobac (1932 – 2021). Spring 2012 with me in Portland OR, Mt. Hood is in the background.

Day 21 remembering my grandfather. Year unknown but likely it’s the mid to late 60s, on Hvar island. Likely new year’s celebration judging by the clothes. With my grandmother Slobodanka.

Day 22 remembering my grandpa. 2017 and 2007, taking the tram to Sljeme. For decades, he would hike up the mountain 3 times a week, until the pandemic started. After the quarantine he had a hard time getting back in shape, but by Sept this year he was back to once a week schedule

Day 23 remembering my grandfather. Fall 2019, just after I got my US citizenship. My goodness, he was so happy for me. A big part of why I wanted to get it was the ability to leave the US for longer periods of time, with the ability to return. But then the pandemic hit.

Day 24 remembering my grandfather. Likely late 60s at a Dinamo Zagreb game, taken by Zvonko Dogan (1921 – 1996), Dinamo’s club photographer and former player. Also a rare photo of my grandpa smoking. He quit decades ago and was militantly anti smoking by the time I was around.

Day 25 remembering my grandfather. Unsurprisingly, it’s yet another New Year’s Eve! December 31st 1964. A few days before my grandfather’s 33rd birthday. My grandma is younger than I am now.

Day 26 remembering my grandfather. 2012 in front of Multnomah Falls in Oregon.

Day 27 remembering my grandfather. 1994 with me in Žumberak.

Day 28 remembering my grandfather. 1993 in Velebit, on the 60th anniversary of the Premužićeva Trail, a 57 km long hiking trail along north and central Velebit. Taken by Vili Strašek (1914 – 1999), a passionate photographer of Velebit.

Day 29 remembering my grandfather. 1980, checking out the first news camera with “integrated sound recording” (his description on the back of photo) at TV Zagreb. I assume this records on VHS tape (looks like an empty case on the desk?). Grandpa is on the right wearing a suit.

Day 30th remembering my grandfather. One of the earliest and most memorable photos. My grandfather in 1932, at 7 months old .

Day 31 remembering my grandfather. At Dinamo FC stadium, 1977.

Day 32 remembering my grandfather. 2006 in Lonjsko polje.

Day 33 remembering my grandfather. 1984 at the Sarajevo Winter Olympics, on Mount Bjelašnica.

Day 34 remembering my grandfather. 1994 and 2015 at Grafičar on Sljeme.

Day 35 remembering my grandfather. Walking around lake Bajer near Fužine in 2017.

Day 36 remembering my grandfather. Having coffee in Zagreb, September 2020.

Day 37 remembering my grandfather. Taken on his favorite hike to Grafičar on Sljeme, December 2010. It’s only fitting to share this one today after I just did the same hike in his memory.

Day 38 remembering my grandfather. Having a coffee with Mandy on Rab. Taking a break from his usual mountainside excursions, he took us to Rab island in the summer of 2017.

Day 39 remembering my grandfather. 1991 with me. What I love about this recording is that he’s trying out a new camcorder and testing out features. He’s holding a mic in his hand and announcing what he’s doing. It’s so sweet.

Day 40 remembering my grandfather. Dubrovnik in 1962.

Day 41 remembering my grandfather. In London in 1982 with Branko Puharić, former TVZ director and RTZ general director from 1982 – 1986.

Day 42 remembering my grandfather. At the Monument to the Revolution of the People of Moslavina in Podgarić.

Day 43 remembering my grandad. 1984 with Ivica Cvitković.

Day 44 remembering my grandpa. This is on Velebit, sometime in the 80s. You can see by his outfit that he’s not an experienced hiker yet. He would later get to know every corner of that mountain, but it’s so inspiring to me that he was already over 50 when he first started hiking

Day 45 remembering my grandpa. Group photo with Tito. I have no idea who this group is and what the exact year is (likely early sixties?), but apparently grandad wasn’t too impressed

Day 46 remembering my grandpa. 1985 at Lake Ohrid.

Day 47 remembering my grandfather. Somewhere on Velebit, April 2003. Resting .

Day 48 remembering my grandfather. 1986 in New Orleans. He always said he enjoyed his time in that city, but I never saw any of these photos until after his passing.

Day 49 remembering my grandfather. At Amsterdam Schiphol airport in 2012.

It’s been 50 days that I’ve been sharing memories of my grandfather. I shared the first one on October 22, a week after he passed away. I don’t know how much longer I’ll be posting these, but for now, a photo from Krk in February 2015, and another one in the same spot 3 weeks ago

Day 51 remembering my grandfather. In Venice in 1954.

Day 52 remembering my grandfather. September 2020 in Zagreb.

Day 53 remembering my grandfather. This is from 1979 and it’s probably a work pic from TV Zagreb.

Day 54 remembering my grandfather. This was taken in October 2019 at our usual coffee spot.

Day 55 remembering my grandfather. A video still from 1993 or 1994, I think.

Day 56 remembering my grandfather. Žumberak, 1996.

Day 57 remembering my grandfather. Taking a break, July 31st 2021. Our last hike together.

Day 58 remembering my grandfather. He took this shadow selfie a few years ago, I found it on his laptop. I love how there is so much of him in this photo.

Day 59 remembering my grandfather. Giving a lecture at Komrčar park on Rab island in 2017.

Day 60 remembering my grandfather. With art school friends from ŠPU in front of Umjetnički paviljon in Zagreb, 1955.

Day 61 remembering my grandfather. In front of Crater Lake visitor center in Oregon. We tried to reach the top, but the road was closed due to snow. I never made it back there

Day 62 remembering my grandfather. In January 2011 at the airport, before my departure to Portland after winter break my freshman year of college.

Day 63 remembering my grandfather. Another shot from the time I left for the US for the first time. August 2010, pouring beer into my college branded nalgene

Day 64 remembering my grandfather. My grandad had a “government birthday” in January 1932, but he always said that date was made up after WWII, because the Italians burned down the church in Lika with the birth records. Instead, he claimed he was born on Christmas Eve 1931, so today he would have been 90 years old

Day 65 remembering deda Looking at cat videos with his granddaughter

Day 66 remembering my grandfather. Hiking on Sljeme, 2010.

Day 67 remembering my grandfather. Lonjsko polje, 2002.

Day 68 remembering my grandfather. With his mother and younger brother in Kutina, just before she passed away, 1964.

Day 69 remembering my grandfather. Screenshot from a video, summer 2017 I think.

70 days remembering my grandfather. Skype screenshot from 2016, showing me a loaf of bread?

Day 71 remembering my grandfather. Celebrating granddaughter @ver4pantera‘s birthday, 14 years ago

Day 72 remembering my grandpa. The last new year celebration before my grandmother Slobodanka passed away, January 1st 1989. She loved new year celebrations (see this thread for more), but this may be the last one he attended? I’m sad he won’t see this year, whatever it brings.

Day 73 remembering my grandfather. Mid to late 50s in Kutina.

Day 74 remembering my grandfather, September 2007. These were taken with a Canon 400D, my first DSLR which he got for me just before this hike. He always supported me in my photography efforts. Feels like I just took these, but it’s been almost 15 years.

Day 75. Another shot of grandpa in Venice in 1954.

Day 76 remembering grandpa. Work ID from the 70s.

Day 77. Today would have been my grandfather’s 90th (official) birthday. He didn’t celebrate birthdays, but when he was (politically) forced to retire in 1991 he said he started counting birthdays backwards. So in this photo from 2013 he would have been 37 by his own count

Day 78 remembering my grandfather.

Day 79 remembering my grandfather. This is Garić grad fort in Moslavačka gora, built in 1256, abandoned in 1544. The photo was taken in 2015.

Day 80 remembering grandpa. Opening a present 5 years ago.

Day 81 remembering my grandpa. The photo in the frame is from 6 years ago. I took this photo of it today, the light through the window was perfect.

Day 82. Here’s the full photo without the crop. He was such a jokester (he’s drinking Velebitsko beer on Velebit mountain)

Day 83 remembering. 1977

Day 84 remembering my grandfather. 1957 during mandatory military service in the JNA in Pljevlja, Montenegro. Afterwards he attended reserve officers’ school in Karlovac and trained as a combat engineer.

Day 85. Zagreb public transit pass, issued in 1980. The photo is probably a bit older. My grandfather loved public transit, trains, buses and trams.

Day 86 remembering my grandfather. In Portland, 2012.

Day 87 remembering my grandfather. This may be at the BBC in London in 1977.

Day 88 remembering my grandfather. Having coffee together in 2017.

Day 89 remembering my grandfather (yesterday’s post came in late). Video call screenshot from September 2020. Showing me roses in the garden

Day 90 remembering my grandfather. I already posted this one in my eulogy, but let’s have it here as well. With my mom and grandma sometime in the mid 1960s (slide photo)

Day 91. Coffee after a hike in August 2017.

92 days remembering my grandfather and today I bring you this great one of him with the World Trade Center in the background.

Day 93 remembering my grandfather. Holding my mom in 1965 I think, slide photo.

Day 94 remembering my grandfather. In front of Trakošćan castle some time in the 1950s. It’s interesting how the castle tower has a cupola in this photo — it must have been removed some time soon afterwards because most photos of Trakošćan show the castle without it.

Day 95 remembering my grandfather. So this one is interesting, it’s labelled “Autoput ZG – LJ 1958.” so it’s from the Brotherhood and Unity Highway youth work action during which 58 000 volunteers built the highway between Zagreb and Ljubljana in 8 months! He may be on the left?

Day 96 remembering my grandfather. With Mandy in Voloder, 2015.

Day 97 remembering my grandfather. Cleaning a fish on Krk, 1994 probably?

Day 98 remembering my grandfather. With my mom in 1964

Remembering my grandfather day 99. Circa 1993.

100 days remembering my grandfather – with my sister 13? years ago

Day 101 remembering my grandfather. In Žumberak in 1994, photo taken by me (definitely one of my first shots)

Remembering my grandfather, day 102, 1958.

I’ve been remembering my grandfather here for over 100 days. I’m concluding this daily ritual for now to create some personal space for closure. I may continue posting photos of him here, just not every day. This is the last photo of us together, in July 2021.

Okay I know I said I won’t be doing daily pictures, but I came across this shot today and just have to share it. My grandpa with his trusted Lada, built like a tank and absolutely peculiar. No power steering but with power windows. The car I learned to drive on.

The post Remembering my grandfather for 100 days appeared first on Vlado Vince.

The socialist Yugoslav state was in many ways an aberration of the polarized Cold War period. Socialist, but not Soviet-aligned; friendly, but not exactly allied with Western Europe and the U.S.; its unusual position produced unique developments in computing. At our current tumultuous historical moment of the pandemic, worsening climate crisis, and most recently the Russian invasion of Ukraine, we may be witnessing another global polarizing moment that may have long term political, cultural, and technological consequences. By looking at unusual technological developments from the late Yugoslav period – the curious case of Iskra Delta and DEC collaboration, the history of the Galaksija (the Yugoslav DIY microcomputer), and the development of JUPAK (Yugoslav Packet Network) – I offer a few lessons as we potentially move into a world where technology is once again an integral part of geopolitical conflict.

The post Revolution During Disintegration: Lessons from a Short History of Yugoslav Computing (HOPE 2022) appeared first on Vlado Vince.

Note: this is the English translation of the obituary I wrote for my grandfather in Croatian in October 2021. You can read the original here.

Josip Gobac, the Director of Television Zagreb from 1978 – 1986, passed away from COVID-19 on October 14th. He was not only a successful journalist, a talented painter, an unstoppable hiker and a big nature lover, Joža Gobac was also my grandfather. He died at the age of 89 after catching the virus that he kept warning us about. Although he was fully vaccinated, he unfortunately couldn’t beat the disease that has taken away thousands of our loved ones in the last two years. On October 14, 2021, 19 more people died from COVID in the Republic of Croatia, and like most of those who have died since the beginning of the pandemic, they were marked in the collective memory and the media only with their initials and year of birth. They all deserve to be remembered. This text is my attempt to remember my grandfather, but also the man who left a huge impact on the development of journalism and media in Croatia and touched the lives of many people.

Josip Gobac was born on January 6, 1932. Or maybe on December 24, 1931. He was born in Lički Osik, or maybe in Mušaluk. Like much else in the life of Joža Gobac, the truth depended on how it was written in official documents, and the story that my grandfather would tell us. One thing is for sure, he was born in the early 1930s in Lika, and then he moved with his family to Popovača, where he spent the first seven years of his life. His father was a coachman with the German Getz-Sever family, so the whole Gobac family – mother Marija, father Nikola, sons Joža and Baja – lived in their big house in Popovača. At the end of the thirties, they moved to Kutina to a small house in the current Ljudevita Posavskog Street.

He grew up in poverty, but his childhood was interrupted by the war and the occupation of Kutina by the ustashas and the Germans. He would speak about this period with extreme clarity – as a child, Joža witnessed the worst downfall of the human race and he told the truth about that time until the end, regardless of how it was forgotten or deliberately distorted over time. Grandpa would tell us about Jewish families who were taken to death camps during the night, but also about the experience of living in a multicultural environment that Kutina was before the war. My grandfather carried vivid memories of the Feldbauer, Freidferd, Strauss and other families who represented the small but significant Jewish community of Kutina and the surrounding area. Joža often retold the same stories, but from time to time he would share a new one. Last year I heard about how as a child he helped neighbors with the Sabbath and Passover service – something he never mentioned before. All these short anecdotes have always been very valuable to me, because my grandfather carried the memories and the experience of a world that is now long gone.

After the war, he was sent to the gymnasium in Krapina, and in 1947 he participated in the youth work action (radna akcija) to build the Šamac – Sarajevo railway. He was always happy to talk about youth work actions, and they were important to him for two reasons: firstly, after a poor childhood and a devastating war, Joža actively participated in the reconstruction and industrialization of a country that was, by his description, “a backwards and illiterate society”. That experience stayed with him forever. Secondly, the work actions showed that Joža was more talented, and probably more interested in writing than using a shovel, so he got his first journalistic assignments by writing “propaganda materials” during work actions. In the following years he continued to write for youth newspapers, and from 1954-56. he was the editor of Moslavački list. He spoke about that period in an interview for Moslavački list no. 1453 in June 2007, which was published in memory of him on the website of the City of Kutina.

My grandfather continued his education at the School of Applied Arts in Zagreb, and he remained close to his alma mater, on the one hand through close friendships with artists such as Miroslav Šutej, whom he would refer to as a “Schulkollega”, and on the other hand through an extremely pronounced artistic sentiment which he combined with clear journalistic expression. That was the last time he moved, this time to Zagreb, with which he had a complicated but very special relationship.

In 1960 he graduated from the Higher Pedagogical School in Zagreb and began his career at Television Zagreb, where he remained until his forced retirement in 1991. At that time Joža met and married Slobodanka Trbojević, a young chemist and later a full professor at the Faculty of Economics in Zagreb. In 1963, Joža became the editor of the School Television, and in 1964 their daughter Ira, my mother, was born.

He considered School Television to be his greatest professional accomplishment. He believed in the educational potential of the media, and TVZ developed a school curriculum that became a key element of the education system. He has also collaborated with other European public televisions, and was a member of the Eurovision Group for the Educational Program.

At that time he was also very active in Dinamo FC. He was a big sports fan, and in the late seventies he was on the expert committee which organized the selection of the team that won the Yugoslav Championship in 1980 and 1982.

Joža had a very active career on television. In 1976 he graduated in pedagogy from the Faculty of Philosophy in Zagreb, and in 1978 he was promoted to the position of Director of Television Zagreb — a position he says he never wanted, but that he completed the job he was assigned. He managed the construction of the TVZ House complex on Prisavlje, a capital project that he pushed through to the end despite extremely unfavorable conditions in a country that was then increasingly suffering from inflation and general economic and political instability.

It was during this period that he first traveled to the US, first to New York, the city where I now live and which he described as too dirty and noisy for him (which is still true!), and then to Grass Valley in California. There he experienced something that left a great impression on him – at one point their driver stopped in the middle of the road with a stop sign, despite the fact that the intersection was completely empty. Joža asked: “so why are you stopping?” The driver pointed to the stop sign and said: “these are the rules!”. He was extremely impressed by this observance of the law, so he often mentioned this anecdote as proof that America is an organized state. Various other examples and my experiences could never convince him that not everything is so great in that country.

After his dismissal from the TVZ Director position in 1986, Joža worked in the documentary program division, and that is when his great love for mountains and mountaineering developed. He became a member of the mountaineering association Industrogradnja and developed a very close relationship with academician and botanist Sergej Forenbacher, a great mountaineer and connoisseur of Velebit. He later made a documentary with Professor Forenbacher, which accompanied prof. Forenbacher’s book Velebit and its flora . He also started making hiking trips to Lonjsko polje, Gorski kotar, Žumberak.

With the disintegration of Yugoslavia, the change of government and the start of the war in Croatia, Joža was forcibly retired in 1991., along with many other “unsuitable” workers and journalists of TVZ, now HRT. Retired too early, Joža dedicated himself to hiking, but also to his family. He first took me to Sljeme in May 1992, before I could walk. We have been hiking together all my life, from that first trip almost 30 years ago, to our last joint hike to Sljeme two and a half months ago.

I always remember my grandfather with a video camera or using his computer, editing some new film. At a relatively late age, he learned to use technology that was new and demanding even for much younger folks, and he was constantly following new trends. In the 1990s and early 2000s, HRT often published his documentaries, which he produced as an outside consultant, and he loved to insert a few seconds of me or our family briefly going through the frame. Over time, HRT had less and less interest in the work of its former journalists, so my grandfather devoted himself to film production in his hometown. Every year he made films for the town of Popovača and for local companies and factories. In the later phase of his life, he dedicated himself to the place of his youth, which has always been very important to him. He also had his films broadcast on Youth Television (OTV), later Jabuka TV.

Although he retired more than 30 years ago, when I would walk around town with my grandfather it would be like walking with a movie star. We would get approached constantly, with folks coming to say hello to him, and he often could not even remember who all these people were, because he was not only popular but also very sociable.

When he came to visit me in 2012 while I was studying in Portland, he often hung out with my friends at the Marino café run by Dario from Banja Luka. Since Joža never learned English (he knew enough to use a computer), he spoke Russian to Russian literature students. Apparently, during the war, Joža learned the basics of Russian from the Cossacks who were stationed in Kutina with the Germans. The Cossacks hung out with the children of Kutina while taking their horses to pasture, and little Joža was taking care of his cow. It took him a bit of time to get comfortable in Russian, but that would be enough to share essential information.

Grandpa combined his love for art, nature and education in a very unique way. Although he sketched and painted only occasionally after he graduated form the School of Applied Arts, he expressed himself artistically in the composition of his shots and in video editing. He was extremely proud of his granddaughter Vera, whom he always claimed to be an exceptional artistic talent, and she found her expression in a very similar way, combining architecture and urbanism in her own work.

In recent years, we talked via video call more often than in person, even before the pandemic. He always asked when my wife Mandy and I will be coming to Croatia, and he would ask us to come up with a plan of what we’re doing together in advance! From day one he loved Mandy so much and he enjoyed showing her our country. We went together to Voloderske jeseni, to Sljeme, to Krk, to Velebit, to Gorski kotar, to Rab. He was very excited to make a printed guide to Dubrovnik for us, but he did not want to come along, because of the crowds.

In early 2020, my grandfather was the first person to warn me that something serious was going on. Being quarantined was extremely difficult for him, especially in the first days when he could not go to Sljeme. Despite this, he was cautious from the start and he followed all public health instructions. He was among the first to be vaccinated in March and he was so happy to hear whenever one of us would get vaccinated. He talked about the importance of following instructions until the end, but unfortunately, like many others, he caught the virus which he tried so hard to avoid.

Before he died, I was full of rage, full of rage at those irresponsible, at the masses of people who deny the existence of a pandemic, who do not wear masks and do not want to be vaccinated. I was most angry at the those who lead death cults and share conspiracies and propaganda, who live on Facebook and in the media, and some of them even inhabit presidential palaces and then manipulate the masses of naive people. But after my vaccinated grandfather passed away, that anger turned into grief, grief for him and all the others who left unnecessarily and prematurely.

I don’t know how to correct the collective mistakes and individual decisions that led to my grandfather’s death. But by sharing his life story, I want to encourage myself to see a valuable life in every death of this mass tragedy. I will carry the memory of his love and care with me forever, as will all of us who knew him. Grandpa lived through hard times and he left us during a very difficult time.

Although I still cannot yet understand that he is no longer with us, I will continue to live in solidarity with others by following his example, believing that in unfair times we must continue with equal care and honesty for all, because that is the only way we can continue.

Rest in peace.

If you knew Josip Gopac and want to share some story about him, or you also lost a person to COVID 19, feel free to contact me at vladimirvince at gmail.com.

I started a YouTube channel with his documentary work and I will be posting new videos as I find them in his sprawling archive. You can check it out here.

The post Remembering Josip Gobac, my grandfather appeared first on Vlado Vince.

The first computer I remember using was my grandparents’ Windows 3.1 machine. This was some time in the 90s, and while it used 5.25″ and 3.5″ floppies as removable storage, it also had a hard drive and for all intents and purposes wasn’t that different from computers we use now. I can literally boot up DOS and Windows 3.1 on a modern laptop without too much trouble right now. The 8-bit computers like the Galaksija were different, and for me one of the most fascinating aspects about them was the idea of storing programs on audio cassettes and on paper!

This wasn’t unique to the Galaksija, of course. All the other major 8-bit machines of the day like the Commodore 64 or the ZX Spectrum also used cassette tape interface for cheap data storage, but most of the other major Western machines soon introduced support for cartridges and floppy disks, or even hard disks, as they were much easier to use and way more reliable. However, since the Galaksija was literally a computer designed around specific import limitations, and people building them had limited spending power in the first place, it relied on the slow and cumbersome, but cheap and accessible audio cassettes.

Cassette tape storage: technical background

Some basics and oversimplified facts first — all digital data consists of bits: units of 1 and 0. In computers, we often count those bits in groups of 8 as bytes. These days we transfer millions of bytes per second when we doom scroll Twitter, but at the end of the day they’re still just ones and zeros. While today those ones and zeros may be going between our phone to a cell tower through the air, or if we’re old school through ethernet cables between our computer to a switch or a router, at the physical level we are still transferring electrical signals of 1 and 0 – on and off.

The same thing happens when using cassette tapes as storage. To store data on a tape, Galaksija generates modulated sounds that corresponds to ones and zeros it can understand. Above is what half a second of that sound looks like in Audacity (this is the code to the Snake game). Below is what that sounds like (warning: check your volume before playing!).

The sound is modulated to achieve appropriate efficiency for the hardware. Tomaž Šolc has a good write up on this, including the comparison between Galaksija and ZX Spectrum’s tapes.

But how exactly does this sound go in/out of Galaksija? Typical stereo analog audio consists of 3 wires: left, right and ground. Galaksija uses one channel for input, one channel for output. You connect it to your tape deck’s headphone jack to load, or the mic jack to save files.

Distributing software for the Galaksija

But how was software for the Galaksija published and distributed? Unless you wrote the program yourself, in the 80s you had three options:
– obtain a listing printed out in a magazine and type it manually
– buy or borrow a cassette tape
– tape a program being broadcast on the radio

It was common to publish programs in Računari and other computer magazines. Here’s an example of a hex loader program published in Svet Kompjutera in February 1986. You would type this in and then save to tape for future use.

There have been some commercial programs (mostly games) sold for the Galaksija, but most software was shared for free in the community. Here’s an example of an educational tape for math, published by Zavod za udžbenike i nastavna sredstva, found and made available online by Marko Šolajić.

From what I can tell, Zavod za udžbenike i nastavna sredstva was the only serious publisher of software for the Galaksija. They solicited submissions from programmers in magazines and then published them on tapes. Here’s an ad from Svet Kompjutera magazine (November 1984) where Zavod za udžbenike advertises a competition for Galaksija, “Sinkler” (I assume this is the ZX Spectrum) and Commodore 64 programmers.

Here’s another interesting Zavod za udžbenike ad (Svet kompjutera 4/84): you can order Galaksija computers, accessories, tapes (chess, math, demo tape) and books and manuals.

The coolest way to get a program was to “download” it over the air. In Belgrade, late Zoran Modli would broadcast Galaksija programs on his radio show Ventilator 202. The listeners would prepare their recorders and then tape the awful sounds coming from the radio! Ventilator 202 was a very popular show, and Zoran Modli was another important part of the emerging community at the intersection of tech and culture. Here’s an interview with him from 1990.

Modli was also a professional pilot and flew 727s for Jat Yugoslav Airlines. He sadly passed away last year, but will appear posthumously in the Galaksija documentary currently in post production.

Getting Galaksija software today

Let’s take a look at some Galaksija software that is available to us to download and load onto our Galaksijas today! There are three types of files you can find online these days:
– scans of text files to type in
– .gpt files that are HEX dumps of BASIC or machine code
– WAV recordings ready to load.

retrospec has one of the best organized collections of Galaksija compatible software, but everything is available in .gpt files, not wav. This is where Tomaž Šolc saves the day once more. His Galaksija development tools feature a gtp2wav program that will convert any .gpt hex dump into a Galaksija compatible wav file! Once you obtain your audio file, you’ll connect the Galaksija tape interface to your sound output. Then you’ll type in OLD command and run it, and then play your file/tape. The screen will turn black while the tape is loading.

At this point you can run the program by typing RUN! BASIC programs can be viewed easily by hitting the STOP/LIST button or typing LIST. Machine code programs will show up only with the BASIC command 10 A=U. (&2C3A) telling the computer to find the program at memory address &2C3A. Many games were written in machine code because it ran much faster than BASIC, allowing for speed needed in games. Here’s what LISTing Tetris looks like. Just one command telling the computer where to find machine code.

But some software is in BASIC. This is Žaboglavac, a game published last year for the Galaksija and the ZX Spectrum by RETRO INFO, a computer heritage organization based in Zagreb.

As comparison, this is what the Comets game looks in hex dump of machine code.

Selection of cool Galaksija programs

Sintesajzer

One of the coolest Galaksija programs I’ve found is Sintesajzer. This program let’s you use your Galaksija as a synth by utilizing the tape output. Galaksija doesn’t have sound support, but it generates tones to transfer data. This program flips that around to generate sound! Here’s what that sounds like in my poor rendition of Jingle Bells. I’m pressing keys G, H, J, K and L. The screen turns off while you play. I think this is for the same reason it turns off while you’re loading/saving to/from tape — to conserve processing power normally used for generating the image on the screen. The program seems to have been recreated/re-released in 2006.

Night Pilot

Another fun game is Night Pilot, a “fly by instruments” game, that lets you fly between the capitals of Yugoslav republics and independent provinces. What I love about this game is that it’s such a great representation of what many games for Galaksija are like – due to limited graphic capabilities, most is left to the imagination. Most of the time you’re just adjusting numbers: roll, pitch, speed, flaps. You have a basic map. The only moment when something shows up on the screen is when the airstrip shows up for you to land. Game description explains why: this is night flying, so lights have to be turned on the runway for you to see it!

The game was published in Svet kompjutera in December 1985. It consists of over 2 pages of assembly code to type in! It also requires ROM B to work, otherwise it will fail on startup.

Inspektor Spiridon

I love this game! You play as Inspektor Spiridon, the world famous detective who needs to collect the keys in the Haunted Castle in order to find a secret code. The game is actually pretty hard (as you can see from my play-through), so it’s no wonder that the author, Dragan Janković, included an add-on cheating program that lets you select infinite lives, invincibility and level selection.

Ziul

No 8-bit computer would be complete without a text adventure, and Galaksija had Ziul, written by Dejan Ristanović. You play as Ziul the dwarf, who finds himself in the mysterious labyrinth of the dangerous wizard Marduk. Marduk is, thankfully, not at home, but many of his evil ghosts are roaming the labyrinth while he’s away. The only way out for Ziul is to find the magical ring that will open an exit.

Superšah

This is a chess program, called Superšah. It’s significant because it was one of the few monetized Galaksija games, originally published by PAGE SOFTWARE and advertised in Moj Mirko in September 1984. The game was then picked up by Zavod za udžbenike which published a slightly different version.

Conclusion

Compared to other 8-bit computers, the variety of software for the Galaksija is very modest, even when compared to some other Yugoslav machines like the Orao. The software available mostly consists of games — there is virtually no productivity software like text editors, or other utilities that may be used in business. But this makes sense, given that the Galaksija was a DIY computer meant for folks who were just starting out and learning about computers and programming. Even though most programmers were amateurs, there was already a move toward professionalization and monetization of software. Ironically, local programmers wrote more professional software for Western computers like the ZX Spectrum than for the Galaksija itself (see Kontrabant and Vruće letovanje). But that shouldn’t take away from the importance Galaksija had as a teaching platform. While only few published games survive, there were likely countless practice programs which the thousands of Galaksija users wrote while teaching themselves BASIC and assembly programming.

This concludes my three part series about the Galaksija. I’ve truly enjoyed researching this wonderful computer and all the amazing people who’ve worked on it and used it. Building a Galaksija prompted me to dive deep into the history of Yugoslav computing, and in addition to researching other Yugoslav microcomputers like the Orao, I’m currently exploring other areas of computing including the mainframe and mini computer eras, but also the development of early networks and BBSs in Yugoslavia. If you have any stories about the Galaksija or other 80s microcomputers that you would like to share, please reach out! You can find me on Twitter or email me at vladimirvince at gmail dot com. I would love to hear your story!

The post Building a Galaksija, Part III: Audio cassettes, magazine listings and radio waves appeared first on Vlado Vince.

Recently I watched Cathode Ray Dude‘s videos on faxing from Windows and dialing up at home, and I was inspired to experiment with modems and faxing myself. He has a great article on configuring an ATA (analog telephone adapter) to place calls between phones/faxes/modems without a need for a telephone network. I went ahead and got myself an Linksys SPA-2102 ATA, but despite the seller claiming it was unlocked, I found that it was still locked to SunRocket, a VoIP provider that used these devices in the mid-2000s.

Below is my documentation on how to unlock a Linksys SPA-2102 ATA if you get a locked one from eBay in the year 2021. These steps are all from different guides dating back to 2007/8, but here I’ll summarize the whole thing.

Linksys SPA2102 is an Analog Telephone Adapter. A device you can use to connect an analog (touch tone) telephony device like a phone, modem or fax, to a VoIP digital network. It is designed to work with a VoIP service like Vonage, but it can also be used locally without it. Since a lot of these devices were sold to customers by VoIP companies in the late 2000s, they were locked to specific service. Many companies went bust very quickly, leaving customers with essentially useless boxes, even though they are compatible with other services. One of these companies was SunRocket, which shut down in 2007 (they later sued Vonage accusing them of stealing their customers). When SunRocket shut down, thousands of customers ended up with locked Linksys SPA2102-R boxes

At the time, there were tons of posts and instructions online on how to unlock these devices, but they mostly relied on SunRocket using one passcode to lock the admin web interface (which had to be unlocked via a phone), and another password to access it.

But all these solutions assumed SunRocket used just a single combination. This wasn’t the case, as they started to change them even before they shut down. So you have tons of people in comments unable to use this “exploit” since their passwords were never set to those values

So, very simply, how is a SPA2102 “locked”? There’s 3 parts. First, the admin web interface is disabled, so you can’t even enter the admin password even if you do know it. You just get a 403 forbidden error trying to go to SPA2102IP/admin

Then you need the actual admin password, to access the admin interface. And finally, you need to prevent the device from getting the configuration that locks it again (this used to be a problem in the 2000s, but it’s unlikely any of the servers are active now, mine wasn’t)

So, if you have the admin menu unlock code and the admin password, this is pretty easy. But what if you don’t? Some folks put together a live Linux CD that was supposed to unlock it for you (I tried it, didn’t work), but you can also upload a new config with the code/pw you know.

So here’s the idea. This ATA has two ethernet ports: one for it’s “downstream” LAN, where it functions as a gateway and DHCP server, and a “WAN” port that goes into your router and provides internet access. The VoIP companies would push configuration updates via this connection.

The IP, path and file name were all hardcoded into these devices. So, if you know that information (and it’s available for the SPA2102), you can set up a server that pretends it’s Sunrocket, and deploy your config with the code and pw to unlock! So here’s what you need:

• an analog phone, or ability to listen to sound via twisted pair and generate dtsm tones
• a dhcp server
• a web server
• the custom config file with a known unlock code and password
• patience

First the config file — back in the late 2000s people were not sharing this openly, but what the hell, here’s the link. Use it responsibly, and verify its safety independently. (mediafire.com/file/ebjrshliv…) This config file does two things: it changes the admin interface unlock code to 79309925, and it sets the admin password to qwe%zxc5

So now that we have the config file, we need our DHCP and web servers. I just configured dhcpd and Apache on a Linux box. If you don’t know how to do it, just look it up, doesn’t make sense to include instructions here. I’ll go over my configurations though. Disclaimer: you don’t need a DHCP server. You could technically set the WAN IP manually via the ATA’s phone IVR menu, but it’s horribly clunky so it’s easier to just get yourself a DHCP server. If you really want to do it, check the IVR codes here: (cisco.com/c/en/us/suppor…)

This is my dhcpd configuration. I set my computer’s NIC to 67.133.234.53 (Sunrocket’s config server’s IP), and I’m issuing IPs on the 67.133.234.53/24 subnet.

Before we move to the next step, you’ll need to get your ATA’s MAC address. It’s written on a sticker. When you get it, rename your config file to macaddress.cfg. So like 718f2e3035ae.cfg (this is random). This is because the ATA will look for a cfg file with it’s own MAC. Now we need to set up a web server (won’t go into how here) and place the cfg to the following path: ataIP/gms/linksys/SPA-2102/2102/718f2e3035ae.cfg (again this is a random MAC, you need the one for your device)

Once we have all this ready, the next step is to connect a phone to the ATA (or in my case a speaker connected to twisted pair going into the device, with alligator clips connected to a 3.5mm audio cable coming out of my computer, with DTSM tone generator website open)

Per the config (which I saw after unlocking), the ATA should look for a new config file (resync) after resets, upgrade attempts, and randomly. For me, it wasn’t doing it after resets at all. But I’ll show you how you can generate a reset.

Before you get started entering the codes via phone, connect your ATAs WAN (“internet”) port to your computer’s NIC that is set with the static IP and DHCP server. Verify that an IP on the correct subnet was issued. With a phone you can control the ATAs IVR menu (link to codes earlier in the thread). Pressing starts the IVR menu. You can enter 73738# for “factory” reset (just reset to cfg defaults). Some folks reported that the ATA would try to download the cfg after this. Mine didn’t. Instead, what worked for me was entering the wrong web enable password, and ignoring the wrong password alert. Here’s how:

• with the phone, enter * * * *
• enter 723646# to access web menu unlock
• enter 79309925# as password (probably you could enter anything tbh)
• press 1#
• press 1#
• hang up

It’s useful to look at your web server access log, since that will show you when the ATA attempts to get the cfg file (the first few requests here were my test from the local machine). You can see those two last GET requests coming from the ATA (which got IP 67.133.234.10)

Here’s another proof of successful download. I was running tcpdump on the NIC, and these are the packets with the GET request. If you don’t know your IP, path and cfg file name, you could do packet capture like this to learn the values for your device.

So now we hopefully have a successful upload of new cfg. We still need to enable the web menu, but before we do, I suggest you unplug the “WAN” cable going to your web server (or rename the file on the web server), so that your ATA doesn’t re-grab the cfg. Here are the admin menu enable steps:

• with the phone, enter * * * *
• enter 723646# to access admin menu enable settings
• enter 79309925# password. It should now be accepted
• enter 1# to confirm
• enter 1 (without #) to save
• hang up (IMPORTANT!)

It’s important to hang up. Since I don’t actually have a phone, my line was “off hook” while the speaker was connected, and I didn’t realize it. So the settings weren’t applied. If you need to put the “phone” “on hook”, just disconnect the twisted pairs from speaker. This diagram was useful lol

Ok! Now you should see the Admin Login link in your Linksys web interface (you can access it via the yellow “Ethernet” port, it will give you an IP and be at 192.168.0.1) and you can login with:

username: admin

pw: qwe%zxc5

You now have complete admin access to your device. You can change SIP setting, Line settings, everything. You may want to disable provisioning to prevent download of cfgs. The server is still online, but didn’t have mine: http://67.133.234.53/gms/linksys/SPA-

To disable provisioning (download of cfgs) change Provision Enable to No.

This is pretty much it. I ended up upgrading the firmware to latest version offered by Cisco (5.2.13(004)) and I used a similar web server process since their windows exe method didn’t work for me: (software.cisco.com/download/home/…)

In that Provisioning menu there is a section for firmware upgrade. It should be set to Upgrade Enable, and you can also set the upgrade URL. Important — you actually need to change the file name to something different than the firmware currently on it, or it won’t run it.

Anyhow, now I can fax and call phones (which I don’t have) and call modems within my home without any analog phone service, which is fun. I could also use this with a real VoIP service technically too. For transparency, this method mostly comes from this great video from 13 years ago. Also some steps from this archived blog post (mainly initiating the cfg download by entering the wrong admin menu password) (docme.su/doc/408123/how…)

The post Unlocking an SPA-2102 ATA in 2021 appeared first on Vlado Vince.

The post Biking in Brooklyn on November 6th 2020 appeared first on Vlado Vince.

In the summer of 2020, a number of stories about Galaksija –the famed “first Yugoslav microcomputer”– came out. There was the Tribune piece that connected the Yugoslav computing industry and the development of Galaksija to the Tito-Stalin split in 1948. A number of articles were published about the Galaksija documentary currently in post-production and the crowdfunding effort around it. These stories, along with a few Twitter threads, piqued my interest. Despite working in tech and years of interest in the history of computing, I realized how little I knew about the Yugoslav domestic computer industry and culture of my (former) country. Hence I embarked on a journey to research this topic by attempting to build one of these early microcomputers, and to read as much documentation and original sources I could find. What follows are personal impressions and notes. Through this process I encountered a number of other fascinating stories that require additional research — this is an early draft of perhaps a larger project exploring Yugoslav tech heritage and its cultural impact.

Most articles about Galaksija start with a bold statement: Galaksija was Yugoslavia’s first computer! While that’s an understandably effective way to get the reader’s attention, it’s plainly false – Galaksija was definitely not the first, and it actually came out quite late. Released in 1984, it was one of later Yugoslav computers. However, this framing is understandable – Galaksija was by far the most successful domestic computer in terms of cultural impact and name recognition. But this framing of “the first” obscures what I think is the really fascinating thing about Galaksija: it was a uniquely anachronistic computer that should not have existed were it not for the specific circumstances of the late Yugoslav socialist state. After months of research, I think that what makes Galaksija unique is how it successfully overcame strict bureaucratic rules through means of technical design, and then it helped change those rules. Combined with an incredibly successful publishing effort in Računari u vašoj kući magazine, in the course of a year Galaksija reached the homes of thousands of Yugoslavs, and educated tens of thousands more about home computing. Moreover, within the same year, the strict customs limitations that prompted its design were relaxed – a significant policy change which Galaksija certainly influenced. Voja Antonić, Galaksija’s designer, makes a similar argument in his autobiographical article Hacking the Digital and Social System.

But let’s take a step back! Institute of Nuclear Sciences Vinča and the Mihajlo Pupin Institute, a University of Belgrade research institute, developed the first Yugoslav digital computer CER-10 (Cifarski Elektronski Računar, “Digital Electronic Computer”) in 1960. CER-10 was a vacuum tube, transistor and electronic relay based mainframe computer. It was initially deployed at the Tanjug News Agency, and was soon followed by a series of CER models released through 1975 and used by state institutions and companies, banks and the military. However, despite the early entry into the computing industry, Yugoslavia abandoned its domestic mainframe development, and by the mid-70s shifted to importing IBM and Honeywell models (Ei-Niš even manufactured Honeywell H6000 mainframes under license). 

In 1971 Intel released the 4040, the first commercially available 4-bit microprocessor, followed by the 8-bit 8080 in 1974. It was soon followed by Motorola 6800, MOS 6502 and Zilog Z80 — by the mid 1970s computers utilizing 8-bit CPUs like Altair 8800 and the Apple I started the home computer era. Despite the early mainframe efforts, it took a while for Yugoslav companies to start producing microcomputers. Unlike the widespread import of mainframe machines from the West, strict customs rules significantly limited access to popular microcomputers like the Sinclair ZX-80 and ZX-81, cheap home computers that were accessible to many Western European families. This is where the story gets murky – and incredibly interesting! For items that couldn’t be obtained at home, shopping abroad was a very common thing for Yugoslavs, especially the middle class. Two popular destinations were Graz in Austria and Trieste in Italy. The practice –allowed and often encouraged by the authorities– was one of the many perks of Yugoslavia’s unique geopolitical position. Unlike the citizens of other European socialist states, Yugoslavs had largely unimpeded access to the West. Shopping abroad made them feel like they were a part of an increasingly global consumer culture, but it also provided opportunity to obtain new technologies like VCRs and camcorders, and most importantly – computers. However, there were limits.

The limits were regulatory, but also financial. You could only import DM 50 (Deutsche Marks) worth of goods, so people smuggled stuff all the time. There were no computers available for less than DM 50, but smuggling was a national sport by this point – you might get a ZX Spectrum along with some Levi’s jeans in Trieste (despite Levi’s literally making jeans in Yugoslavia, but I digress). Or perhaps a relative working abroad in Germany might bring one. While smuggling was easier for Yugoslav citizens closer to the Western borders with Austria and Italy, computers were still largely inaccessible to most Yugoslavs. Additionally, it was difficult to learn about computers in the first place. Until 1984 there were no computer magazines – early hobbyists would have to get publications like the American BYTE magazine the same way they would get their computers – from abroad. 

It’s commonly assumed that the Yugoslav authorities didn’t allow import of microcomputers in order to protect the domestic microcomputer industry. Perhaps a more likely explanation is that by the early 80s the Yugoslav economic crisis and the significant bureaucratic and institutional rot affected the state’s ability to adapt to new circumstances the same way it did in earlier years. It would far exceed the scope of this project to analyze economic policies and collapse of the Socialist Federal Republic of Yugoslavia, but this is a good intro if you’re interested. After a few decades of huge improvement in standard of living, Yugoslavs experienced rapid stagnation and reduction by the 1980s. This was particularly manifested in inflation which would reach record heights in the late 80s, and “FR Yugoslavia” (Serbia and Montenegro) would trump even those in the 1990s. However, for a relatively stable socialist economy until then, inflation of the early 80s was hugely disruptive.

By 1984, when Galaksija was released, a few Yugoslav companies dabbled in production of microcomputers, mainly for use by state institutions and companies, or for education. Some of the computers that came out around the same time as Galaksija were:

• Lola 8, an Intel 8085A based computer made by Ivo Lola Ribar, a manufacturer of heavy machine tools named after a Partisan hero
• Galeb and Orao, developed by Miroslav Kocijan and manufactured by PEL Varaždin, 
• Partner, manufactured by Iskra Delta
• Ivel Ultra, an Apple II clone designed by Branimir Makanec and manufactured by Ivasim elektronika
• Pecom 32/64, manufactured by Ei Niš

None of these achieved particularly widespread personal use – they were prohibitively expensive for individuals to purchase and were often produced in small quantities. Since they weren’t widespread in home use, there was limited software, and no community. This is something that took me a while to understand – at the time, most computers were not compatible with each other, so software had to be developed specifically for every model. Communities formed around specific machines. While none of these computers achieved the widespread recognition Galaksija had, the circumstances of their manufacturing are fascinating: they were often produced by companies that had no previous connection to computing. Famously, to manufacture the Orao, PEL Varaždin produced wicker baskets for export to offset the cost of importing parts for their computer (another import regulation intended to keep import deficit low and bring foreign currency into the country). 

Galaksija’s origin story is legendary at this point: Voja Antonić, an engineer from Belgrade, came up with the idea for a DIY microcomputer while on honeymoon in Montenegro. This wouldn’t be his first computer – he had already  designed another one by this point – the EL-82 for Elektronika inženjering. Around the same time, the 21 year old Dejan Ristanović was working on a special edition of Galaksija popular science magazine which would focus solely on computers. Računari u vašoj kući (Computers in your home), the special edition of Galaksija magazine, was supposed to come out in the fall of 1983. However, in early summer Zoran Vasiljević introduced Dejan and Voja, and they decided to combine their efforts: Voja would complete the design of the (then still unnamed) DIY computer, and Dejan would include the instructions in Računari u vašoj kući. 

The project quickly exceeded this limited scope. They worked through the fall to finalize the computer, writing the miniscule 4KB ROM capable of running the computer, generating video on the Z80A CPU and featuring a very limited BASIC interpreter. The PCB was single sided, requiring a total of 119 wire jumpers, but lowering the cost significantly and making it simple enough to produce at home if necessary. Finally, the Z80A CPU ran at 3.072 MHz and the board featured 3 sockets for 6116 static memory, allowing for a total 6KB of RAM (expandable up to 54KB with an extension board). Space was made on the board for a second EPROM chip, so Galaksija was designed to be software upgradable. 

But Računari u vašoj kući went a step further – in addition to providing instructions, schematics and documentation necessary to build this computer, they also organized a supply and distribution chain! In the October 1983 issue of Galaksija magazine they introduced the project and released a questionnaire asking their readers if they would like to order a DIY kit. They expected a few hundred requests at most, but even this preliminary questionnaire exceeded 1000 orders (You can find this questionnaire on page 62 of Galaksija issue 138).

So, how does a magazine organize a supply and distribution chain for an affordable DIY computer in a country with strict customs rules? Let’s take a look at how a Računari reader in early 1984 could get their hands on components necessary to build a Galaksija: 

• The “mechanical components”: PCB, keyswitches and keycaps. Institut za vakuumsku tehniku from Ljubljana made the keyswitches, Mipro and Elektronika from Buje made the PCB and the rest. You would order these parts directly through Računari for 5600 dinars. 
• The ICs: these include the Z80A CPU, 7400 series logic ICs, RF modulator, the crystal and 3 IC sockets. Sold through Microtechnica in Graz, Austria for 1660 Austrian schillings. Ordering was coordinated so that no single package exceeded 1500 dinars. The instructions are fascinating: you would first write a letter to Microtechnica, they would send you an invoice (this is all through mail), you would then go to a bank to send the money, and then wait. You could pay by Amex, Diners, Eurocard or Visa. I’m not sure how widespread these cards were.
• The EPROMs: Galaksija used 2732 and 2716 ROMs to store the OS and characters. Per the ordering instructions, folks who ordered the ICs from Microtechnica would receive their EPROMs already programmed. It’s unclear if they were included in that 1660 schilling price. But the EPROMs would come from Belgrade, where Voja Antonić programmed them. Računari offered to program EPROMs for folks who obtained them independently completely for free! They would just need to send them in with a pre-stamped envelope.
• Passive components: there was no mention of ordering more common electronics parts like capacitors and resistors. I assume that those were available locally in stores.

For folks who weren’t ready for DIY, and for use in schools, Elektronika inženjering and Zavod za udžbenike i nastavna sredstva made pre-built Galaksijas. These weren’t ready yet when Računari came out, but eventually many schools received these factory built Galaksijas.

The decision to hold off on publishing Računari in the fall, and instead to work with Voja to include the Galaksija DIY guide was incredibly important for Galaksija’s success, but it also helped start the Yugoslav computer publishing industry. Before Računari there were random articles in newspapers and magazines. By 1985 there were computer monthlies like Moj mikro and Svet kompjutera,

Dejan’s story really resonates with me, because he is someone who both understood computers but also how important they would become. He also understood and excelled in communication – publishing that first edition of Računari reached tens, if not hundreds of thousands of readers. In the late 80s he founded one of the earliest and biggest bulletin board systems in Yugoslavia – Sezam BBS. He wrote for Računari u vašoj kući for many years, and then started PC Press magazine. Dejan’s website is a great resource on all things Galaksija, Računari and other 80s computer projects, as well as his later work.

Računari u vašoj kući issue 1 hit the newsstands in late December 1983 and the initial release of 30 000 copies quickly sold out. Eventually, over 120 000 copies were sold after 4 re-prints. Instead of a few hundred, over 8000 people ordered the kit or sent their EPROMs in to be programmed. Galaksija was immensely popular, and even the people who didn’t end up making one had an opportunity to learn computing basics. Ristanović included a series of articles on choosing your first computer, with reviews, specs, benchmarks and comparisons of the popular 8-bit machines of the time: ZX Spectrum, Commodore 64, Apple II. Even the IBM PC and PCjr were included – the future was just around the corner, but for now all the attention was on Galaksija. The issue included an article explaining how computers worked and what they could be used for – at the time people still didn’t know what uses computers had at home. A large section was focused on games – the original home computing killer app. Along with the Galaksija DIY guide –which included the schematics, parts list, step-by-step instructions and even the guide for making your own RF modulator and logic probe– there was also an extensive intro to programming in BASIC. While a number of programs were eventually released for Galaksija (mostly games), for many people it was their first opportunity to experiment with programming – first in Galaksija’s simple BASIC, and later on in the Z80 machine language.

The rest is history. Even though the delivery of kits suffered many logistics issues –including the global static RAM price increase of 1984– by the end of the year thousands of households had a Galaksija. Računari u vašoj kući soon split from Galaksija magazine into an independent computer monthly, and other magazines like Svet kompjutera followed the development of Galaksija software and hardware expansions. But ironically, Galaksija was a victim of its own success. By late 1984 the SIV (Federal Executive Council) passed a new customs regulation that significantly relaxed the customs limit and explicitly allowed import of microcomputers. Per the February 1985 issue of Računari u vašoj kući, the regulation still didn’t allow mail ordering of complete computer systems and many machines like the Commodore 64 were still too expensive according to the rule, but popular ZX-81s and ZX Spectrums could now be freely bought in Graz or Trieste and brought home without additional anxiety at the border. 

By the late 1980s Yugoslavia was rapidly disintegrating, and the time of the 8-bit home computer was over – the IBM PC, presented as just another option among many in the original Računari computer comparison, was quickly becoming the personal computer standard. However, Galaksija played a unique and important role in popularizing computing and raising a whole generation of programmers and IT professionals. While it was mostly forgotten for almost 30 years, Galaksija has seen a huge resurgence in interest recently. In Tribune, Michael Eby speculates that Galaksija resonates so much with us today because it “embodies a destratification of today’s technological hierarchy, a tacit ideological assertion that computing machinery should be for the masses, cheap and available to everyone, and that neither money nor technical know-how need be barriers to entry”. While this claim certainly resonates with me, after researching this history in detail I have to admit that looking back we may be projecting some of these qualities on to this wonderful little computer. On the one hand, it is true — Galaksija was designed and distributed to the masses by Voja, Dejan and many others involved in the process. The care, solidarity and absolute lack of financial interest in this immensely successful project seems completely incompatible with our understanding of the tech industry in 2020. But the particular role Galaksija played –getting around bureaucratic rules of a socialist state and popularizing computing– ultimately negated these qualities – as soon as these rules were relaxed, the market got quickly dominated by Western produced models. 

However, maybe this kind of projection is exactly what we need right now. Going through the process of building a Galaksija and researching its history helped me develop a completely different way of relating to a computer. By following these 40 year old DIY instructions I developed a uniquely personal relationship to this tiny board of chips and wires I put together. My Galaksija is truly a personal computer — it’s different from all the other ones, and I know it inside out (even though it took me an embarrassingly long time to get it working). While I don’t think we should all go back to making our own computers (although passing Right to Repair laws should be a priority!), I do wonder if there are other ways we can lay claim to the technologies we use – to transcend the cycle of perpetual consumption and waste generation imposed on us through planned obsolescence. For me, learning this history is the first step.

The post Building a Galaksija, Part II: The Origin Story appeared first on Vlado Vince.

At the start of the coronavirus pandemic, I unexpectedly took up an old hobby of mine – repairing old electronics. It started in April, with an older, but still relatively modern Macbook. Soon I tackled a 1993 IBM Model M keyboard, and then around the end of summer I attempted my first serious retro build: the original Galaksija 8-bit Yugoslav microcomputer, designed by Voja Antonić in 1983. It was all thanks to my first visit to the PEEK&POKE computer museum in Rijeka, Croatia.

I was completely amazed by this museum, but despite the extensive collection on the ground floor which includes the IBM PC, Macs of all types, Apple IIs and NeXT workstations, I got the most excited when I climbed to the second floor and discovered all the domestic, Yugoslav built computers. Growing up in the 90s and the early 2000s, I never got to experience the period of 8-bit microcomputers, and in school I never learned about domestic computers like the Galeb, Orao or Galaksija. It was so exciting to see them all in one spot, so this visit inspired me to research this topic more. As I started learning more about this period, I came across a number of fantastic projects, like Josip Perušanec’s Orao and Galeb emulators and Orao and Galaksija clones. Tomaž Kac’s website was amazing: hundreds of old Yugoslav computer magazine scans, including program listings, schematics and other stories from the 80s. I made a decision: in order to better understand this old technology, but also the tech culture which is still relevant, I will build one of these computers from scratch.

Thanks to lots of available documentation and the fact that it was originally designed for DIY, I decided to build a Galaksija. Right around this time a crowdfunding campaign was started to fund a documentary film about Galaksija. The campaign also included an option to order a new DIY kit. I thought about ordering one, but I decided to construct it on my own: the kits are scheduled for delivery in 2021 and I wanted to get started right away. Nonetheless, I still supported the crowdfunding effort and I’m looking forward to seeing the movie some time next year! This ended up increasing my cost somewhat, since I ended up ordering quite a few wrong parts that had to be replaced, but I like to think that this helped me learn more! Regardless, I truly enjoyed the process of building a Galaksija, and I’m excited to share my story.

Just like all my projects, I primarily documented it on Twitter. I started with a few self imposed rules:

• I will try to use the original or identical parts as you would in the 80s
• I will primarily used the original DIY guide from the first issue of Računari u vašoj kući magazine (full magazine scan available at retrospec.sgn.net). I will use other sources only if absolutely necessary.
• I will build the original Galaksija with 8KB of ROM (ROMs A and B) and 6KB of static memory
• I won’t build my own power adapter, instead I’ll use an off the shelf 5V/2A (the original guide included instructions on building your own)

I started by making a spreadsheet and finding equivalent parts to those listed on page 54 of Računari u vašoj kući issue 1.

I managed to find most parts on Mouser Electronics. There were a few exceptions:

• 2 x 2732 EPROM and 1 x 2716 EPROM: got them on eBay (I ended up getting two pairs of 2732 due to issues with my EPROM programmer)
• 6116 RAM: I originally got it on Mouser, but it was the wrong width. I got replacements on eBay
• Z80A CPU: originally I got a completely wrong CPU on Mouser. Then I got a new old stock 1983 SGS Z80 on eBay
• Keyboard switches: got them on KBDfans

Along with these mandatory components, I got an EPROM programmer and eraser. The programmer I got (LAQIYA TL866Ⅱ Plus) ended up being the worst possible choice, but more on that later.

The most important component was the PCB. The schematic was published in Računari, but making a DIY PCB is beyond my current capabilities. Moreover, most DIY Galaksija builders in the 80s got their PCBs through Računari, so I decided to order my PCB as well. Luckily, in 2020 it’s fairly easy and cheap to order a PCB online, and thanks to the great gerber files hosted on spetsialist-mx.ru “Russian Galaksija website”, there was no need to convert the old magazine scan. I got a pretty good price on jlpcb.com, but the minimum order was 5 boards. I decided to get them anyway – I ended up sending 2 to Croatia and one to Macedonia, so eventually we may have 4 new Galaksija from this single order! I used the 5th board to experiment with making a case.

When I got the PCB, I started by soldering 119 jumpers. While reading the guide in Računari I already realized that this was going to be tedious work – if only I took the warning seriously enough! I didn’t count them carefully enough and I missed two crucial ones. I only discovered this later on, but more on this later. I used copper wires from CAT 5 cables, as a sort of an homage to the original guide which suggested using telephone wire. After completing the jumpers, I got started with the IC sockets. I made a few mistakes when ordering these (mainly wrong widths), but I also messed up with bad quality sockets for the memory chips. Once the sockets were in, I soldered the capacitors and resistors. After soldering 117 (!) jumpers, these were a breeze.

Then I moved onto the keyboard. I had some issues with the key switches too: due to the difference in connector placement on modern key switches vs. the PCB designed for those from the 80s, I couldn’t solder them directly. I ended up adding some extra copper wire in order to make the connections. This sort of worked, but it isn’t ideal. Luckily I don’t have any issues typing, but some keys are a bit unstable. Coincidentally, Voja Antonić, Galaksija’s original designer, just published a new version of the PCB on Hackaday that features new key switch connections, including a few other details (mainly the double layer PCB that doesn’t require jumpers).

Then I got the EPROMs and the EPROM programmer! I had no issues with the 2716 2KB EPROM which is used as a character generator ROM by Galaksija (it stores the graphics to use for character generation). However, my LAQIYA TL866Ⅱ Plus continuously failed to burn the 2732 EPROMs used for ROM A and ROM B. After a lot of trial and error, I realized that the 2732s I got needed to be programmed with 21V, and my programmer tops out at 18. I managed to resolve this by connecting the programming pin to a 21V power supply, but I still couldn’t successfully burn it. At most I could get the first out of every 8 bytes. I won’t go into too much detail here, but you can read the whole story on Twitter.

In the end I destroyed one 2732 and programmed the other one with the help of 32 spreadsheets and 32 separate burn attempts. This way I got ROM A ready, and I got a 2732B EPROM for ROM B. 2732Bs are programmed with 12V, but I still had issues. It turned out 2732s require 50ms programming pulse, which the TL866II Plus again couldn’t deliver. If you’re choosing an EPROM programmer for 2732s steer away from this one. In the end I put together a bash script that successfully programmed the 4KB by attempting the burn approximately 1000 times.

My struggle with EPROMs was incredibly frustrating, but I ended up learning a lot not only about EPROMs but also about the actual ROMs used by Galaksija. I learned the most from the complete ROM A disassembly by Tomaž Šolc. This is an incredible document that I’m still using for research. ROM A features video generation functions (Galaksija uses the CPU to generate the image) and essential BASIC commands. ROM B was released in 1984 as the first Galaksija software upgrade, featuring advanced BASIC functions, assembler, support for the monitoring program, printer support and a few other things. I have the copies of the ROMs and some additional documentation on my GitHub.

Once ROM A was programmed, I got the 6116 ROM: in half the width! I ended up improvising a horrible looking adapter, but it ended up being useless since the sockets on the board weren’t good enough anyway (I didn’t know this yet). At this point I finally installed the CPU and realized that I didn’t get a Z80A, but a completely random and unrelated 40 pin model. How I managed to make this mistake – I have no idea. I had to go back to eBay.

When I got the replacement – an Italian Z80 from 1983 – I got excited hoping I’ll finally get some kind of image on the screen. It just wasn’t meant to be. By using a logic probe I found an issue with the CD4017 decade counter: it turned out it wasn’t a CD4017 at all but CD40175BEE4 flip flop with an almost identical name. I had all these issues due to my lack of experience. I started to fill out my Galaksija schematic with notes as I tried to resolve problems.

When I finally got the replacement CD4017, I first got excited and then disappointed again. I finally got some kind of image using my RCA/HDMI adapter which I’m using with my modern LCD. I got image synchronization and the adapter switched to PAL, but other than that the image was completely white. I started reading all the articles I could find about image problems with Galaksija. I related to the article “I’ll kill myself, it won’t work” from Računari issue 3… But I had no solution. Since my Z80 is very similar to a CPU Tomaž Šolc found in an old Galaksija which had image generation problems, I convinced myself that I had the same issue. The original owner constructed a small circuit to synchronize the 74LS166 shift register with the CPU. I ended up making the same circuit – with no success getting an image. It didn’t help since I had no issues with CPU timing, but making this circuit helped me better understand how Galaksija generates image, and how analog image works in the first place. 

In the meantime I ordered a small oscilloscope (YEAPOOK ADS5012h) and started to analyze the signals between the CPU, ROM and memory. Very quickly I noticed that the WAIT and IORQ lines have no signal. But I still couldn’t figure out why. After a long period of troubleshooting, I tracked it down to two missing jumpers at the top of the board. I felt incredibly silly because in order to find the missing jumpers I spent so much time analyzing the signals instead of just counting the jumpers or simply looking at the board. But still – with the jumpers soldered I finally got some characters on the screen!

It’s hard to explain how happy I was to see that first READY on the screen! I had double rows, missing pixels and horrible contrast, Galaksija couldn’t run even the simplest PRINT command, but after all the troubles I finally got something on the screen. In terms of fixing image problems, I fixed the contrast and flickering fairly easily – I experimented with resistors R9 and R10, I lowered the overall power supply voltage from 5V to 4.63V, and I connected the RCA/HDMI adapter to the same voltage. However, I still had missing pixels, double rows and I couldn’t run any commands.

Finally, I accidentally noticed that if I touched the first 6116 memory chip (I got the correct width ones from eBay in the meantime) the image would stabilize. This ended up being my final eureka moment – it turned out that my low quality IC sockets couldn’t provide a good connection between the RAM and the rest of the board. I quickly took out the sockets and soldered the memory directly to the board. I got a clear image and the commands immediately started to work!

When I finally got Galaksija working, I started to try out programs and games. Tomaž Kac’s website had tons of programs with great documentation. I also found many programs in a wav program dump by Marko Šolajić. I had a great time playing Žaboglavac, a game released by Retroinfo in 2020!

To load programs I used my laptop since I don’t have a tape recorder. At first I had some issues due to sound volume, but once I removed the R14 resistor all the issues went away. Over the next few days I had fun with Galaksija and I tried out more or less every program and game available for download online. I then decided to finish this project with three mini projects:

• HDMI: I integrated the RCA/HDMI adapter with the rest of the board
• 3D printed case: I designed a case in Sketchup and 3D printed it
• dump2gtp: I wrote a mini tool to backup Galaksija programs recorded in wav to the popular gtp virtual tape format. The tool is called dump2gtp since it uses MAME to dump audio files. I’m still working on it and the final version should convert wav to gtp directly.

Documentation for these mini projects and all the files are available on my GitHub.

HDMI

Since I don’t have an analog TV or monitor, I’m using a small RCA/HDMI adapter to connect my Galaksija to a modern LCD display. I noticed that the image is much better if the adapter is connected to the same power supply like the rest of the computer. I took out the adapter board from its case, drilled two holes in Galaksija’s board (with a lot of caution not to drill through any connections) and directly soldered analog video to it. Later on I also added a switch so I can use either analog RCA or HDMI video.

3D printed case

I wanted to make a case like the one from the DIY guide. The guide suggested using a blank PCB board as a cheap solution that can literally be soldered together. I even experimented with one of my spare PCBs, but I just wasn’t confident enough in my fiberglass cutting skills. Instead, I decided to go with a slightly more modern option that would still look very similar to the original. I made a 3D model in SketchUp and ordered a 3D print through craftcloud3d. For the bottom I used a plexiglass board. I combined the logos of MIPRO, Elektronika inženjering and Zavod za udžbenike i nastavna sredstva, the three companies involved in Galaksija production, and printed it on hard paper as a label. I’m pretty happy with the final product, and I’ve published the SketchUp file (with a few improvements based on my test print) on my GitHub.

dump2gtp tool

After I finally had Galaksija working, I wanted to be able to save the programs I write on the actual Galaksija as gtp files. Gtp is a popular format used by some Galaksija emulators, mainly MAME. It’s a simple file format that saves Galaksija files as hex bins instead of audio. Many programs available online are only available in gtp files and they need to be converted to wav to be used on real Galaksijas. Luckily, Tomaž Šolc wrote the gtp2wav tool which is available alongside his other great Galaksija development tools. However, I had no luck finding a tool that would do the opposite – convert a wav file to gtp. Such a tool would be very useful to archive old programs found on tapes, but also for saving new programs written on Galaksijas. Since I couldn’t find such a tool, I decided to make one. My tool uses MAME memory dumps and converts them to gtp. I’ve decided to use MAME as a stopgap since I still don’t know how to convert Galaksija’s audio signals into a hex dump directly. The process is:

• You save a program to a computer as wav
• You load it into MAME running in debug mode
• You make a memory dump from MAME into a txt file
• You use my dump2gtp tool to make a gtp file
• You verify the gtp file in MAME

This program is still in a very early development stage. Eventually I would like to make it a direct wav2gtp tool, without the need for MAME. It was fun for me to make since I learned a lot about the gtp format and how Galaksija saves programs on tape through reverse engineering the gtp format (since I couldn’t find any gtp documentation online). dump2gtp is available on my GitHub.

Bonus

Working on this project, I got very interested in small details, like the two different READY logos you can find on different Galaksijas. I’m pretty sure that Tomaž Šolc was the first to document this. Most character generator ROM copies you can find online only feature the Elektronika inženjering logo, so I made a few changes to the ROM file and made a version with the MIPRO logo too. Now you can easily choose which one to burn to your 2716 EPROM. Both versions are available on my GitHub.

The post Building a Galaksija, Part I: The Tech appeared first on Vlado Vince.

Earlier this year, I tangentially followed the story on Western Digital shipping SMR (shingled magnetic recording) drives labelled as “Red” NAS drives. While these drives do provide satisfactory performance for non NAS use, they are significantly slower for most NAS’s, and virtually unusable for ZFS, used in FreeNAS. However, since I don’t really deal much with storage these days, I didn’t actually follow the details. When articles started coming up noting that WD is changing their branding to make it clear which drives are SMR, I assumed the issue was resolved.

Fast forward to September 2020. I returned to Brooklyn from Croatia and found out that one of the 4TB drives in my home FreeNAS server failed. The drives are from 2016, so the warranty expired. I promptly reordered what I assumed was an identical drive. I should have heeded the advice from a few folks on Twitter regarding checking for SMR in model names. I could swear that I did. But it turns out I actually did not… This post documents my process of finding this out the hard way, in case anyone else stumbles upon the same experience.

I ordered an WD Red 4TB NAS drive on Newegg. I can now clearly see that it is WD40EFAX (the SMR version of the same drive I had), and not WD40EFRX (regular CMR drive that WD sold for years), which is now branded WD Red Plus. Since the drive looked virtually the same as the one I was replacing, I was sure I got WD40EFRX.

So I followed the instructions to replace a drive in FreeNAS and… Resilvering failed. When running zpool status -v each time I tried I got a bunch of WRITE and CKSUM errors, with too many errors on the side.

It looked exactly the same as the issue that took my original drive out of commission, so I made three wrong guesses:

• I didn’t “burn-in” the drive
• Mini SAS cable between LSI SAS 2008 and the drives was faulty
• I got a faulty drive

The third assumption was correct, but not in the way I thought, obviously. After I attempted a burn in by running a long SMART test (smartctl -t long /dev/da4 in my case), I still experienced issues. I also tried replacing the cable. Resilver failed each time at ~7%. I should have immediately guessed the problem was something else when the SMART test didn’t come up with any errors, but alas. I RMA’d the drive through WD.

WD RMA experience wasn’t great. I couldn’t do it through their web form and instead had to contact support. Then they had issues with their payment system. I ended up paying $5 for shipping. All of this was unnecessary, of course, since they shipped me yet another SMR WD40EFAX drive.

I basically went through the same experience with the second drive, until I finally checked the model number. Of course, it was the SMR model. I ordered a replacement drive. This time I got WD40EFRX, the regular CMR version now branded WD Red Plus 4TB NAS.

I did come across a great article explaining the science behind shingled drives, which also suggested that they could in fact function with ZFS if write caching and lookahead were turned off. This of course would slow down the resilvering process so much that the drive would effectively be unusable. Since I did have some extra time before the new drive came in, I decided to test it. I couldn’t get FreeNAS to turn off these two functions. I tried adding the following lines to /boot/loader.conf:

kern.cam.ada.4.write_cache=0
kern.cam.ada.4.read_ahead=0

But it had no effect. Instead, I turned them off using a Linux VM running on the same server. I run FreeNAS as a VM on ESXI, and it connects to the LSI SAS 2008 RAID controller through PCI passthrough. This way I could easily access the drive through another VM. I issued these two commands:

hdparm -W0 /dev/sda – turns off write caching

hdparm -A0 /dev/sda – turns off lookahead

This time it worked. This was the otput of camcontrol identify da4 -v in FreeNAS:

And then the long resilver started. It took 5 days to reach 15%, but there were no errors (previously resilver would fail after 7%). At this point my new drive arrived, so I stopped the resilver.

The lesson of this long and unnecessary conundrum is to pay close attention to details and  LISTEN TO PEOPLE ON TWITTER  (well, maybe sometimes). But it also shows what a mess WD did with their SMR naming mixup.

The post Encountering WD SMR mess in real life with TrueNAS/FreeNAS appeared first on Vlado Vince.

In 1948, Yugoslavia was expelled from Cominform, the Soviet information agency, in retaliation for its "non-aligned" status; deprived of information-processing capacity, the country created its own IT industry from scratch.

1/ pic.twitter.com/8PIVqSg24f

— Cory Doctorow #BLM (@doctorow) August 2, 2020

A couple of weeks ago I came across this amazing thread and the recently published @jacobinmag article about Galaksija, Yugoslavia’s DIY computer from the 80s. I was vaguely aware of its history, but it has largely been forgotten in the Balkans. I’ve been on a bit of an old hardware binge lately, so I attempted to source a Galaksija or another locally built computer while I’m in Croatia. No luck so far on the second hand market as it seems like these are increasingly rare. Ditto for the magazine that made them popular.

Growing up in the 90s, I was unaware of the independent computer industry of our former country — the first computer in my household was a PC running Windows 95, and in school they didn’t talk about our socialist IT heritage. However, I then randomly and unexpectedly ended up at @PEEKANDPOKE computer museum in Rijeka, and found not one, but two Galaksijas and a whole bunch of other locally built 80s computers!

I vaguely remember hearing about Peek and Poke many years ago, but somehow I never went there. And then as we were walking around Rijeka we came across this unassuming store front

Turns out there is a treasure trove of old tech here. IBM PCs, Macs of all types, Apple IIs, NeXT workstations… Very cool stuff, though common enough in the US.

There’s a great collection of consoles and games, and you can actually play and use a bunch of machines (after disinfecting your hands)! I especially loved the ZX Spectrum and Commodore 64, popular imported computers for those who could afford them.

But the real good stuff is on the floor above, where you’re welcomed by a display of boxes from the “Factory of computer machines Zagreb” and a large collection of domestic tech industry: computers, calculators, cash machines, phones, radios

Let’s look at Galaksija again. Voja Antonić designed the Galaksija around the Z80 CPU, then readily available in electronics stores throughout Yugoslavia. The schematics and instructions were then published in the special edition of the SF magazine “Galaksija”, hence the name.

The special issue, titled Računari u vašoj kući (“Computers in your home”) went through multiple prints, eventually exceeding 120 000 copies. The hobbyists would assemble the board and then build a case themselves, like the Apple I and other early personal computers.

Fully assembled, Galaksija is a beautiful machine. This one was signed by its maker!

But Galaksija wasn’t the only domestically built computer. There was also Galeb (Seagull) and Orao (Eagle), designed by Miroslav Kocijan for PEL Varaždin and built around MOS 6502 CPU, found in the Apple II, C64, BBC Micro

Here’s the Galeb prototype board

Unfortunately, it took PEL Varaždin 3 years to bring Galeb to market in 1984, by which point it was completely outdated. With only 250 units sold, it was retired and replaced by the faster Orao, which gained popularity in schools as a cheap machine for teaching BASIC.

There are a bunch of other Yugo computers: Delta Parner from Slovenia, the Apple II compatible Ivel Ultra, Pecom 32 from Serbia

Then there’s the calculators! The people of the small medieval town Buje in Istria are proud of their invention — Digitron, the first handheld calculator developed in Europe in 1971!

Digitrons were so popular that even today people here refer to calculators as digitrons… It was such a hit that even Tito had to check it out.

Lastly, you can check out the timeless and impeccably designed Iskra phones made in Slovenia. I remember using those rotary dial ones as a kid.

There’s a bunch of other super interesting hardware made beyond the borders of Yugoslavia, like the bizarre Sun JavaStation “network computer” made to run only Java apps… In 2020 we’re still waiting for Java’s end of life.

But the most wonderful thing about this museum is that it tells the local computer history. Most of these were used by people around here, so many keyboards feature manually scribbled BCS diacritics: šđčćž

ThinkPad 701 Butterfly keyboard from Vlado Vince on Vimeo.

Ahh we need to go back to this amazing technology!

This concludes the tour of the amazing @PEEKANDPOKE museum. There is much more to see, and this small exhibition space showcases just a fraction of their collection. I’m so thankful for their stewardship of our people’s tech heritage, and hope to come back soon!

Make sure to check them out at their website, and look at their hardware donation page — they are looking for some specific pieces, so if you’re want to get rid of something, they may be looking for it! And yes, I did get an Orao mug, how could I not.

I then discovered an emulator for Galeb and Orao computers! I ran the Orao emulator in an XP virtual machine and now I can speak BASIC! I love how the emulator uses the BCS keyboard layout with diacritics even though the Windows locale is set to en-US. Like the real thing!

Both emulators were developed by Josip Perušanec
and can be downloaded here.

I also ran the Orao demo program, loaded from a virtual tape

What is that logo you ask? Well, only the royal coat of arms of Varaždin, given to the city in 1464 by Matthias Corvinus, King of Hungary and Croatia, King of Bohemia and Duke of Austria. Galeb and Orao were developed by PEL Varaždin corporation.

Per the demo program, Orao features a “professional keyboard with 61 keys, split into 3 sections with YU symbols, automatic key repeat and sound (!) on keypress”

Does this mean that typing on an Orao sounded sort of like my cursed beepy Model M speaker mod?? There’s also a web based Orao emulator built by Nenad Mihailović, so you can easily try it. Type BC to start BASIC, choose a program on the side, then type LMEM”” to load the program.

Nenad also hosts the original Orao manual, but looks like it’s been OCR’d so some detail may be missing. His emulator project and Orao page is really well documented though. 

The post Exploring old computer tech from Yugoslavia appeared first on Vlado Vince.

I ended up writing a Medium article documenting my experience and summarizing the procedure, phone numbers, where to go to check-in, etc. However, COVID numbers continued to rise by mid August, and I found myself repeatedly editing this article with manual updates, as new testing locations started opening up. Despite the increased need for testing, official information was still sparse and confusing, but I realized that a static page won’t work for providing accurate and up to date information.

There was also an increasing need for testing information for the rest of the country, so I decided to put together a website that would show all PCR testing locations in Croatia, with correct info like hours, phone, price… At this point the government finally published an official list of testing locations, which I used as the starting point for a very basic, Google Sheets “database” that I published as a json file. I then built a simple website using Umap, an OpenStreetMap project that let’s you build your own maps using geojson.

I was lucky to get help from many volunteers who continue to submit information! I was also joined by Lee Barguss, a Digital Designer based in Rijeka, who made the product easy to use from a UX and design perspective.
We released the website in less than 48 hours and immediately received incredibly encouraging and positive attention from the Croatian media. I spoke to N1 TV about the project and the broader effort of citizens providing mutual aid during the pandemic and coming together to help each other.

You can read more about the project in my Medium post (in Croatian). More tech info is available on Github (where the live version of the website is hosted!), and we post all information updates on our Twitter profile.

The post How I traveled internationally during the pandemic and ended up making a COVID testing website appeared first on Vlado Vince.

As I was researching the history of this keyboard, I found out that per ps-2.kev009.com, these speakers were added to Model Ms meant for use with IBM RS/6000 workstations. I couldn’t find a recording of how the speaker was used when connected to an RS/6000, but per a few forum posts, these were meant to emulate terminal beeping sounds. Unfortunately none of the forum posters reported whether they were able to enable the speaker.

I did find out that the when used with an RS/6000, the speaker is wired to pins F and A on the SDL connector used by the Model M:

I wanted to enable the speaker, so I attempted to wire the spare pins 5 and 6 on the PS/2 side of the keyboard cable. Unfortunately, the pins 5 and 6 on the PS/2 side and pins F and A on the SDL side were not terminated on my cable, making this impossible. While I found some SDL connectors for sale, I don’t have a crimping tool capable of terminating them, so I gave up on this idea.

Instead, I decided to wire the speaker directly, using a 3.5mm audio cable:

The speaker is connected to a line out on my Dell D6000 docking station.

Once I completed the hardware portion of this project, I started looking for a good typing sound emulation solution. I came across the bucklespring program by Ico Doornekamp. Funily enough, this program is meant to emulate the bucklespring sounds of a Model M when using a regular keyboard. Since I already hear the real bucklespring sounds, I decided to modify it to produce beeps instead. The bucklespring program provided a lot more complexity, since Ico included precise buckling sounds for each key, but it was otherwise exactly what I was looking for, including additional functionality such as muting and unmuting.

To make it produce beeping sounds instead of buckle springs, I recorded a 1khz beep using the speaker-test program:

$ ( speaker-test -t sine -f 1000 )& pid=$! ; sleep 0.1s ; kill -9 $pid

Since the original code replicate both pressing and releasing a key (just like a real bucklespring), I removed all the wav files ending in -0 (for release), and replaced all the wav files ending in -1 (for pressing the key). This should be done more ellegantly by changing the code, but for now it works with intended effect.

Additional changes I’ve made:

• defined the model M scroll lock key as 0x77 to use as mute key
• added LED control for engaging the mute key to light up the scroll lock LED (by issuing “xset led 3” to the system)

I run the program with the mute function (-M) and then double press the scroll lock to enable beeping (turning scroll lock LED on). When done with beeping, I double click the scroll lock (turning scroll lock LED off).

Here’s a sample recording, including unmuting and muting the program:

This was a pretty silly project and I don’t expect I’ll be using the beeping sound much, but it was fun to get this speaker working, even in this very work around sort of way. If you know more about how this speaker was actually used (or if you have a recording!), please @ me on twitter (@mejs) or email me at vladimirvince at gmail.

Additional details on the bucklespring program modifications are available on my github.

The post Beepy Model M speaker mod appeared first on Vlado Vince.

At my day job our team manages a lot of laptops, some of which are Macbooks. Since the Macs account for a fairly small share, we never implemented an MDM solution like Jamf, so we largely depend on using scripts to deploy software, and we manually adjust settings. As a mostly Microsoft shop, we use on premise Active Directory, and our Macs use LDAP to bind to the AD domain. Apple’s LDAP implementation is extremely clunky in 2020, and from what I can tell, hasn’t been updated in well over 10 years. After the upgrade to High Sierra in 2017 we even had an issue with LDAP accounts getting completely locked out.

This setup isn’t ideal. Recent versions of macOS struggle to show the “Other users” prompt to enter AD credentials at login screen, and sometimes we have to wait for a few minutes until we can login a new user. We’re also unable to leverage these credentials for 802.1x authentication on wired or wireless networks. And lastly, password sync is incredibly slow as well.

Despite this, it’s still worth it as we often re-deploy laptops and it saves admin time. The Macs are set to allow administration by domain admins, but we also configure a local admin account for backup. Most of our staff doesn’t have local admin access, and with many applications getting updates from the App Store now, this works fairly well. That is — until the user actually needs admin access for something. During normal times, we often provide this in person or over screen share. During COVID times though, it’s all remote.

As we transitioned to WFH, things worked reasonably well for a few months, until we received a ticket about updating some Adobe software, something fairly routine. However, once we remoted in, our local admin password did not work. Nor did the 20 other combinations we tried in case it was initially misspelled. To make things worse, none of the domain admins logged in to this machine previously, so the local admin was the only account with admin access. While there used to be a few ways to recover OSX passwords back in the day, they require physical access to the machine, and file access to the OS drive. Now that macOS uses FileVault disk encryption by default, I don’t think these solutions work anymore. For regular home users, they can use Apple ID to reset the password, but our backup admin wasn’t configured with one.

Now, we wanted to get into this machine without having to wipe it, as reinstalling everything remotely would take a while. Since the machine was bound to our AD domain, our first attempt was to log in with a domain admin account through the terminal. We could log in while connected via VPN, but we were not in the sudoers file, so no admin access was available. A quick Google search suggested that Apple adds LDAP users to sudoers during login process. So the next step was to try to login at the login screen as a domain admin. However, we soon realized that macOS can’t maintain a VPN connection if you return to the login screen, not even if you’re just attempting to switch users. So I had another idea — what if we establish a VPN connection outside the computer, and then bridge it to the Mac? This would be fairly simple on my home network, but way too involved for one of our staff. However, what if we built a small VPN gateway and shipped it to our colleague?

Here’s what the plan was:

• Get a Raspberry Pi and 2 USB ethernet adapters (since new Macbooks don’t have Ethernet any more) + 2 ethernet cables
• Set up an OpenVPN server on premises (our firewall’s vpn client doesn’t support ARM architecture so it won’t work on a Raspberry Pi — learned this the hard way)
• Install Raspbian on Raspberry Pi, install openvpn and isc-dhcp-server
• Use built in Ethernet as WAN. Leave as default (DHCP)
• Configure static IP for USB Ethernet adapter, and enable DHCP server on that adapter
• Configure DHCP server on an obscure subnet so it doesn’t conflict with the user’s home network nor our corporate network
• Configure routing and NAT between “WAN port” (built in ethernet) and “LAN port” (USB Ethernet)
• Configure OpenVPN and set it to auto start
• Configure firewall rules on premises to allow traffic and NAT from user’s home connection

This plan seemed so simple and foolproof, and I quickly tested it using my 2006 black Macbook. This was the critical mistake, as we would soon find out — this laptop did not have FileVault enabled. Everything worked without issues! I connected the “WAN port” to my local network, the “LAN port” from the USB adapter to the Macbook (didn’t even need the second USB adapter since in 2006 Apple still had ethernet on their laptops). Voila, I had the keys to the kingdom. I quickly packed up the whole thing and mailed it to my colleague.

A few days later I am speaking to them over video. They’ve properly connected the RPi to their router and the Macbook. They get the “Other users” prompt. I give them temporary admin credentials to type in. macOS loads. And then they get the SecureToken prompt to unlock FileVault:

This was it. Despite the LDAP config allowing administration to AD domain admins, the latest versions of macOS still require an existing SecureToken admin to create a mobile account and (I assume) add them as an administrator. If the user select “Bypass”, the LDAP user will be created, but it won’t have admin access. We eventually just wiped the machine, since we already wasted a few days trying to make this solution work.

FileVault has (reportedly) been marked as default on Macs since macOS 10.10 (2014), so presumably this solution wouldn’t have worked on most Macs in recent years (if they have FileVault enabled, of course).

This was a disappointing end to what seemed like a fun and quick solution to something that wouldn’t have been a problem back when we all worked in offices together.

The post (Failed) attempt at remotely re-gaining LDAP admin access to a Macbook in 2020 appeared first on Vlado Vince.

I got this keyboard at @lesecologyctr Gowanus E-waste Warehouse. This incredible space in Brooklyn has been a drop off site for e waste, a ReUse store and a Prop Library used by movie and TV productions in NYC for all the historical computers, radios, phones etc

Like the municipal e-waste program which had to shut down due to COVID budget cuts (#DefundNYPD!), Gowanus warehouse is closed until further notice. Before they fully closed, they hosted an appointment only sale and I was lucky enough to make it to one of their last days in May. Somehow I’ve never been to this incredible place despite passing right next to it for years every day on my way back home from work, so I am so thankful to the wonderful staff who let me extend my 30 min appointment to almost 3 hours.

There’s all kinds of old things for sale, from computers, cameras, phones… to sewing machines! And everything is categorized and you can tell that the objects are cared for with love. This is much more than a random thrift store!

I was so tempted to take a Macintosh Plus or Classic home with me, and Nicole, the Reuse Technician on duty, encouraged me to take one with blown capacitors and fix it myself. She normally gets them all in working order, but didn’t get to fix them before the close out.

I got to test a few, but I couldn’t carry one on my bike. It’s so wonderful to be able to physically experience these things. The classic Mac is such a quintessential computer, but they were quite rare in Yugoslavia so they are not something you would just come across

So I compromised and got the IBM Model M and a Kodak Six-20 Brownie Model C. $5 for the keyboard and $20 for the camera. The Brownie was one of the first mass produced cameras (initially released in 1900), and this one may be form 1946. Hopefully it gets its own thread!

Back to the Model M! This keyboard defined the computer keyboard layout that we pretty much use today. Up until 1985, manufacturers would choose their own layouts and add keys as needed. Only the letters and numbers were standardized based on QWERTY typewriter layout from 1870s

So in 1983 IBM formed a group to research keyboard best practices and come up with a new standard. The result was the “101-key IBM Enhanced Keyboard” which included 4 distinctive key areas: typing, numpad, cursor/screen control, and function keys

Another innovation was the buckling spring actuator responsible for its distinctive clicking sound and feedback, reminiscent of typing on a typewriter. These springs are amazing, and still function without issues on my 27 year old keyboard.

Now, most of this stuff was new to me. I knew about the Model M, but I’ve learned most of the stuff I shared here from @benjedwards article on why he still uses his 34 year old Model M (howtogeek.com/661430/why-i-s…)

I really enjoyed this article, both for its historical and technical detail, but also for emphasizing how objects like the Model M represent an important connection to the past. And it’s still fully functional in 2020! It reminds me of an old bike you can easily repair. So back to my Model M. First I had to figure out how to connect it to a modern computer. It comes with a PS/2 cable, ubiquitous up until ~15 years ago, but I couldn’t find any device in my apartment that still uses it. It was developed for the IBM PS/2 system, now long forgotten.

Growing up I had a PS/2 keyboard that came with a little green USB adapter, so I actually made a PS/2 to USB cable, before I realized that this won’t work since the Model M doesn’t speak USB! This should have been obvious since USB was first released in 1996.

So I order an active PS/2 to USB adapter which essentially runs a little board that reads PS/2 and then connects to your computer as USB. So the computer doesn’t know it’s getting the key presses from the Model M, it’s really just aware of the adapter.

Now, at this point I’m impatient and I want to find out if the keyboard is working, so I try to connect the Model M to a Raspberry Pi single-board computer through GPIO pins. Someone wrote a Linux driver for it in 2016, so I attempted to follow their steps (hackaday.io/project/9922-p…)

This… Sort of worked. I did manage to register some keystrokes, but PS/2 requires a clock line between the keyboard and the computer. The driver author got around not needing the clock line by compiling a script that calculates the keyboard clock period.

Which is great, except you do this by holding the “;” key, which was actually broken on my keyboard, as it turns out. Now, I didn’t know this and just assumed that something’s wrong with my wiring, or driver compilation or something completely different. I gave up. So a couple of days later I get the active USB adapter! And I find out that a bunch of keys don’t work… Including “;” . I also notice that there is a pattern, and after checking the connection matrix I’m pretty sure that there are one or two breaks responsible.

So I finally open up the keyboard. I lose an hour trying to get in using my regular interchangeable nut driver, but the adapter is too wide. Luckily the hardware store is now open so this $5 5.5mm nut driver does the trick

Another obstacle! The Model M is so sturdy because it includes the 0.6 kg steel backplate. This is great, except that IBM bolted the rest of the frame to it using plastic rivets, meaning that in order to repair the keyboard you have to break them… So I do, I break them.

Finally I get to the membrane! This was quite innovative in 1985 and most keyboards still use the same system. There are two layers on top of each other, with conductive circles under each key. When a key is pressed, contact is made between circles and the circuit is closed.

As you can see in the previous tweet, someone spilled something on this keyboard at some point between 1993 and 2020. 4 lines were busted, one on each layer. I cannibalized a crappy keyboard I don’t use to try to bridge the gaps. It sort of worked.

Except now I realize that the plastic frame holding the buckling springs is cracked in multiple places, and breaking the rivets didn’t help. In order to replace the rivets I’ll need to do a “nut and bolt conversion” which will place even more stress on this 27 year old plastic..

I could have tried gluing the plastic, but since my membrane fix seemed precarious at best, I decided to order a replacement membrane and frame. Amazingly, these keyboards and parts are still made by @UnicompKeyboard, company formed by ex IBM staff in 1996 based out of Kentucky!

Unfortunately they still make the frames using plastic rivets, and while I tried my hand at riveting using a kitchen lighter (do not do this), I ended up doing the nut and bolt conversion. It took me a couple of tries and this was the best guide: (imgur.com/a/QroSL)

But alas after many frustrating hours I manage to get it just right and get all the keys working. I ended up covering about 80% of rivet holes since I bought out complete stock of M2 screws at Brooklyn home depot (I know they are horrible and I regret it, I went to Lowes first)

So finally it’s time to put everything back together and start clicking! This took a while since I wanted to make sure every key worked well. Here you can see that every key comes with two parts: the key itself and a keycap with the symbol, so it was easy to convert it regionally

Here’s the Model M fully assembled. My regular keyboard is behind it, so you can see what an absolute monster this thing is.

So here’s what typing sounds like. It’s quite pleasant if you’re into it, and you always know when you’ve actually pressed a key. It’s not very pleasant for others in your household though.

Here’s a final shot of the keyboard and the label with the manufacture date. I’m only slightly older!

So that’s it for today. If you’ve enjoyed this, you might like my recent post about repairing a black 2006 Macbook I found in the trash which led me to find out that our corrupt leaders suspended e-waste recycling. Don’t forget to #DefundThePolice!

And while you’re at it, support @lesecologyctr and sign a petition to save curbside composting in NYC. E-waste suspension is a part of the same budget cuts that stopped ALL residential composting in NYC redirecting 59,000 tons of food scraps to landfills (change.org/p/mayor-bill-d…)

There is a separate petition for community composting programs. Programs affected include @lesecologyctr and @GrowNYC which runs Greenmarket composting, which we depended on since we never got municipal composting in our building (bigreuse.nationbuilder.com)

The post Restoring a 1993 IBM Model M keyboard appeared first on Vlado Vince.

We’ve been at home in Brooklyn for two and a half months now. We’ve been lucky and privileged to stay healthy and employed through this crisis. Among other things, as a form of coping I’ve turned to an old and slightly pointless passion of mine: repairing outdated tech.

I used to do this a lot while in college. Get a broken device and fix it, try to sell it on ebay (unsuccesfully). It was fun, but required lots of time and space and money, so hardly compatible with living in NYC. And eventually I ended up in tech so stopped doing it altogether. Fast forward to a few weeks ago. I’m throwing out recycling, and see a box full of laptop parts near the trash chute. A Thinkpad, a Dell and a black Macbook. Screens ripped off. Someone tried hard to make these unusable. Basically an aftermath of the Office Space printer scene.

I tell myself no, I will not get into this. Drop off recycling and go back to apartment. But I can’t resist and go back, just to take a peek at that Macbook. I’m a primarily a PC/Linux user, but that black Macbook always looked so cool and special. Ughh.

So I go and pick it up. The screen’s toast, and the keyboard’s ripped out but seems intact. No HD or RAM or battery of course. I don’t have a charger so can’t check if the motherboard is working. Here we go

A few days later and $40 down I have a replacement screen, 2GB of DDR2 memory and a MagSafe charger. Throw in an old 2.5″ drive from an external HDD and the thing seems to boot up. Need an OS, and it’s missing a whole bunch of screws hence the rubber band on the right.

Now, Apple put the Snow Leopard ISO on @internetarchive so I get it working pretty quickly. I find out the keyboard/touchpad aren’t working because the little ribbon cable connecting them to do motherboard is bust. Maybe that’s why the previous owner was so angry at this machine?

Anyway, as I’m sourcing a replacement ribbon cable and battery, I’m thinking about the rest of the laptops and how they should be properly recycled, even if they’re unfixable. So I go check what’s the deal with e-waste during the pandemic.. SUSPENDED (amny.com/coronavirus/co…)

During normal times, NYC hosts seasonal SAFE events where you can safely dispose of e-waste and other dangerous trash. They also do scheduled pickups. They assume everyone drives to them, but otherwise they are great events.

@NYCSanitation suspended all e-waste recycling until at least June 30, 2021! This is with all the other horrific budget cuts like cancelling summer camps, #SYEP, composting, education and health cuts, basically everything other than @NYPDnews

The impact of these cuts is hard to comprehend. Just the composting suspension will redirect 59,000 tons of residential scraps to landfills, and it affects both community (e.g. Greenmarket) and mandatory composting. Sign a petition to save it here: (bigreuse.nationbuilder.com)

Meanwhile, the efforts to save e-waste recycling are getting much less attention. In the middle of all this, @lesecologyctr was forced to close down their Gowanus e-waste warehouse after 9 years. They are having an appointment only sale through June: (lesecologycenter.org/electronics-sa…)

So I made the though call and didn’t grab the rest of those laptops as there is nothing I can responsibly do with them. At the moment it is both illegal to throw out electronics with trash in NYC, and impossible to recycle them unless you’re a business with $$ to spend. Back to the 2006 black Macbook… After getting it running, I was thinking of what to do with it and figured this might be a good chance to pair it with my 3rd gen iPod from 2003! I got this from a friend in 2008 and never got to use it with a Macbook, so, why not?

Now, an interesting thing about these is they primarily used FireWire to charge and transfer music. They came with two cables connected to a single port, one USB, one FW. So you could connect it to a charger with FW, and USB to computer. Or just FireWire to computer.

I was excited to charge it and transfer music using a single FW cable! I never got to do this when I used this iPod with a PC. However, no luck. I had to upgrade iTunes to get it to work, and Apple disabled FW support on later versions (discussions.apple.com/thread/3802750)

The FW port is sort of useless for an iPod in 2020. Interestingly, the iPod does seem to detect it and shows the “Do not disconnect” message, but the computer doesn’t mount it or indicate it’s connected in any way.

So I come up with another challenge. Can we get this 2014 version of iTunes to download a podcast and then sync it to this 2003 iPod? It’s how the podcasts got their name, so it has to work, right!? A bit of personal history. My first intro to podcasting happened in 2007 when I attended a podcasting workshop in Zagreb hosted by @MaMa_Zagreb and led by @bicyclemark and @timpritlove. Podcasting was such a strange and DIY thing back then!

@bicyclemark is one of those rare podcasting pioneers who is still on air after almost 17 years. I remember listening to his amazing stories from around the world and feeling like I’m getting to experience something unique and special. His dispatches still have that quality! Anyway, back to 2006 Macbooks and 2003 iPods. Podcasts are, or rather, podcasts used to be just audio files published with an RSS feed, that wonderful web 2.0 technology that brought us blogs and news and was then promptly obliterated by Facebook and Twitter

Now, many podcasts still have an RSS feed somewhere, even though so many are moving to centralized platforms like Spotify. And of course @bicyclemark has an RSS feed right on the website! Here’s what those look in Chrome which doesn’t read RSS anymore (citizenreporter.org)

So, even though by 2014 (the version of iTunes I installed) Apple already had their podcasting walled garden, they still had a legacy function of manually subscribing to a podcast by pasting their RSS URL… And it still works!

And here we go. With just the RSS feed I downloaded the latest episode of @bicyclemark‘s show. He’s been putting out many episodes during the pandemic and I feel so privileged to hear from him and dispatches from his guests from around the world.

Now, why am I downloading the podcast onto an iPod? Well, the way you used to do this whole thing is you would sync your iPod every day, or every few days, and the computer would download the latest episodes for you! Hence — podcasting

In any case, I’m now going to enjoy this episode of @citizenreporter and for those of you who’ve made it to the end, I hope you’ve enjoyed this journey through my chaotic pandemic electronics repair hobby. (citizenreporter.org/2020/05/tony-p…)

The post Repairing a 2006 black Macbook during quarantine in NYC appeared first on Vlado Vince.

The post Biking in NYC, May 2020 appeared first on Vlado Vince.

2019 was an incredibly difficult year for cycling in New York City. 29 people riding bikes had been killed, almost three times as many as in all of 2018. We have lost so many dear people to senseless vehicular violence which shows no sign of slowing down. Meanwhile, the lack of response from the people responsible for ensuring safe infrastructure and traffic enforcement has been deafening. Ever since the killing of Robyn Hightman and the Washington Square Park die-in protest that followed, I’ve been feeling increasingly unsafe on a bicycle and have often opted not to ride instead. I’ve heard the same thing from many others, and some folks have stopped riding bikes altogether in fear of becoming yet another statistic.

The one refuge I’ve found over these last few months has been Mechanical Gardens, NYC’s only bike co-op. It is an amazing place, a community bike repair shop located in a church basement in Williamsburg. We welcome everyone and help them fix their bikes. It sounds pretty simple, but it is so much more. It’s a place with a focus on inclusiveness and solidarity, not just as something given, but as actions that need to be performed every single time someone new steps through the door. In the short time I’ve been volunteering with the co-op I’ve learned so much about teaching others with respect and care.

Wrenching at Mechanical Gardens has been a good way to focus my energy on something positive, and getting to know people who volunteer their time or drop in to get help has been incredibly rewarding. It was through Mechanical Gardens that I met Josh Bisker who originally introduced me to the bike co-op at the 2018 TransportationCamp NYC. Soon, Josh and I realized we had something else in common other than our love for teaching people how to repair bikes: both of us have a parent living with Multiple Sclerosis. Josh has been planning to do Bike MS, the National MS Society’s annual fundraiser ride, for many years and I jumped at the opportunity to join him! We started a Mechanical Gardens fundraiser which raised $3660!. There was only one problem — I had already booked a long overdue trip home in October, so I wouldn’t be able to join him for the 100 mile ride. We were eager to do the ride, so a decision was made: Josh would ride for his mom at BikeMS NYC up the 9W road in New Jersey, and I would ride for my dad independently in Croatia, joined by my partner Mandy.

Multiple Sclerosis

What is Multiple Sclerosis? MS is an unpredictable, disabling disease that destroys parts of the brain and central nervous system, disrupting the flow of information within the brain and to the body. The results are both cognitive and physical: extreme fatigue, problems with memory and concentration, blurred vision, loss of balance, dwindling mobility, and more. MS is different for everyone, and it’s still a fairly new medical field. That makes it all the more challenging to solve, but it also means that every dollar leads to new research that makes a serious difference in people’s lives. 

The route

But where should we ride to? We wanted to do at least 100km, but the route had to be close to the city and preferably accessible by train so that we could get home easily in case of any bigger issues. We looked into doing a ride through Zagorje, the hilly region just North of Zagreb, but we got discouraged by the temporary closure of the local train line and the difficulty of getting out of the city without it. After some internet and Strava research, we eventually decided on Sisak – a smaller town just South East of our city.

The Bike

We didn’t travel to Croatia with our bikes, even though we really wanted to try flying with Bromptons. A friend of mine lent us a mid-2000s Scott hybrid for Mandy to ride, while I took my dad’s old Rog Senior. This bike was made in Slovenia and my dad rode it in the 80s. The bike then spent the next 20 years in a shed until I started restoring it about 10 years ago. I’ve added a Brooks B17 seat, switched out the wheels and tires, replaced the unfortunate cottered cranks and installed aero brake levers and modern side pull brakes. Finally, in preparation for this ride, I installed a new stem I got at Mechanical Gardens. It was lovely to have this bike connected to the co-op now! Rog bicycles have been made in Ljubljana since 1949. Rog made all sorts of bicycles, but their most famous model was Pony — a compact folding bike based on the Italian Graziella. They are still ubiquitous all over former Yugoslavia. Rog still makes bikes and their old factory has been transformed into a squat in Ljubljana, providing a space for community organizing, art and various other activities. It sounds similar to a co-op.

The ride

We left the house at 7:30am. It was a Sunday and the streets were completely empty. Even though it was late October, the temperature was unusually warm, the way October days have been feeling more like summer with each year. We took the almost direct route out of the city through Heinzelova and Radnička streets, starting at Kvatrić square. This route follows a wide avenue, but it includes a comfortable protected bike lane (technically located on the sidewalk but with good delineation and lots of space) and it goes on uninterrupted for 8km! While the route is comfortable, the street itself is not very pleasant. There are 6 car lanes with pretty fast traffic and almost no pedestrians — it mostly consists of decaying factory buildings that are slowly getting replaced with mid rise offices. The median includes a transit corridor for streetcar tracks which may get built one day (Zagreb has a large streetcar system but has built almost no new infrastructure in the last 30 years).

We soon reached the Domovinski bridge over the Sava river. The Sava goes through Ljubljana, then passes Zagreb and finally flows into the Danube in Belgrade — connecting three capitals of former Yugoslavia. The bike lane goes over the bridge, but the surface is full of bizarre craters — I’ve never seen this on a paved road before and it’s been like this for years. The bridge is followed by another gap in bike infrastructure. As it is about to cross the A3 motorway, the bike route disappears and we are redirected for almost 4km to a smaller overpass because the larger interchange doesn’t include access for bikes. This is extremely uncomfortable and it’s not marked at all — luckily I’ve ridden here before and knew how to cross the big motorway.

After the unpleasant detour we reach the bike route again and cross the new road to Zagreb airport (D430) which surprisingly also includes a protected bike lane! It’s a shame you have to navigate the secret overpass to get there. It was time to leave big roads and bike lanes behind. We crossed the D30, which loops around to Velika Gorica, and got onto county road 3069, heading East.

It’s crazy how quickly our surroundings changed! Within minutes we were surrounded by fields and passing through small villages. The road was narrow and there was barely any traffic.

The villages we passed through were small and picturesque. Early on we came upon a stork’s nest on a utility pole. Unfortunately, the storks had already left to Africa for the winter.

After a quick coffee break in Obed (population 51), we continued on South, now closely following the Sava river. After a little while, there was a turn towards the river and we figured we’d check out the bridge… To our surprise, there was no bridge and instead the crossing was served by a small cable ferry! Before the city expanded, ferries were common in Zagreb as well. The last one closed down in 1959.

As much as we wanted to cross the river and explore the Eastern shore of the Sava, we were on a tight schedule, and the cable ferry was on the wrong side. We continued South following the flood embankment on the West side. The narrow river road meanders through small villages, picturesque but uncomfortably empty. Many of the houses we passed by were old but well maintained. Others were boarded up and obviously empty. Curiously and quite wonderfully, very few cars passed us, and the people we came across either walked or rode bikes. We soon noticed that most of the people we passed were either very old, or very young.

In one of the towns by the river, Desno Trebarjevo (Trebarjevo on the right, across from Lijevo Trebarjevo, which is on the “left” side of the river, naturally), we noticed a monument to two men in front of a house. The monument honors Radić brothers who founded the Croatian Peasant Party (HSS), a major political party in the early 20th century. HSS is still around and currently holds 5 seats in the Croatian Parliament. Radić brothers are prominently featured in the Croatian education system, yet I never realized they were born in the small village of Desno Trebarjevo, population 396.

We passed other monuments too. Among many small churches and Catholic shrines, there was a significant number of monuments to the Narodnooslobodilačka borba (NOB) — the National Liberation Struggle of Yugoslav partisans during World War 2 against Nazis and local collaborators. During World War 2, this area was under control of the Nazi puppet state Nezavisna Država Hrvatska (NDH) which operated concentration camps, including Sisak children’s concentration camp (officially called “Shelter for Children Refugees”). It was a part of the larger Jasenovac concentration camp complex where thousands of Serbs, Jews, Roma and other victims were systematically murdered. After World War 2, Yugoslavia built monuments to the victims of fascism and partisan fighters all over the country, but during the violent disintegration of the Yugoslav federation in the 1990s, many of those monuments were purposefully destroyed or abandoned. The monuments we came across, however, were kept in reasonably good shape. They were small monuments, not the internet famous abstract “Spomeniks” that often appear on social media, drawing comparisons to science fiction and aliens. Their presence is a reminder of difficult and complex history, both in terms of atrocities they honor, and the more recent conflict that largely sought to erase them.

Biking through this beautiful region is incredibly conflicting. The beauty of nature is contrasted to the lack of people. We were less than an hour away from Zagreb, the economic and population center of Croatia, yet it felt like the only people still living here were the ones too old to move away. Croatia has been losing population rapidly, especially since it joined the EU. And I’m personally a part of this large group of younger people who no longer live in this country, coming back once or twice a year at most.

As we passed through one empty village after another, I could also not stop thinking about the many refugees less than 100km away in Bosnia, trying to cross into Croatia and continue West towards the rich EU countries like Germany. While most aren’t interested in staying here, the Croatian authorities have been violently and illegally pushing them back. “Push back” is the euphemistic term for illegally deporting people suspected of crossing the border without authorization. Croatian police catch them, beat them, steal their money and destroy their cell phones before they are transported to the border with Bosnia and forced to walk back through forests and mountains. These deportations are not documented, and the victims are not given opportunity to claim asylum. For years, Croatian authorities denied the practice of push back, until the former President Kolinda Grabar-Kitarović openly confirmed it on Swiss TV last summer. The combined absurdity of witnessing empty villages and knowing that the police is “protecting” the border with violence is painful as much as it is absurd.

After some 3 and a half hours of riding, we enter Sisak. Sisak is not a large city (population: 50 000), so our first indication that we are now in the city is a simple welcome sign. Sisak is an ancient town. It was originally settled by the Illyrians and the Celts as early as 4th century B.C. and was then known as Segestica. The Romans attempted to take it over on few occasions before 35 B.C. when the Roman legions, led by Gaius Octavius (later known as Augustus, the first emperor of the Roman Empire), captured the town and founded the Siscia military camp, which soon took over the whole settlement. Siscia became the capital of the Pannonia Savia province, and its city mint produced coins that were used all over the Roman Empire. In the 16th century, Sisak was part of the Habsburg Empire and site of 4 large battles between the Ottoman Empire and the Habsburg Monarchy. While Sisak was occupied by the Ottomans on at least one occasion (after the 1593 battle), it was soon taken back by the Habsburgs and started to develop rapidly following the wind down of Ottoman wars in Europe in the early 18th century.

Sisak developed into a commercial and industrial center, first on the two rivers Kupa and Sava, and later on the railroad between Sisak and Zidani Most (Slovenia) which opened in 1862 as the first railroad in Croatia. In the 19th and early 20th century, Sisak specialized in iron works and oil industries. On June 22, 1941 the first detachment of the Yugoslav Partisans was established in the Brezovica forest near Sisak, starting the anti fascist resistance movement which eventually liberated Yugoslavia from the Nazi occupation. After the war, rapid industrialization of Sisak continued under socialism, with prominent metallurgic, oil, chemical and wood industries. At this point the population was rising rapidly, reaching 60 000 by 1991.

During the 1990s war in Croatia, Sisak was 10km away from the front line in Petrinja, and experienced shelling and mortar attacks from the Yugoslav People’s Army (JNA) and insurgent Serbs. These attacks targeted industrial infrastructure, but they also damaged residential buildings and killed civilians. At the same time, 109 Serb civilians in Sisak were killed and disappeared during 1991 and 1992. Two former Croatian police officials involved in the disappearances were convicted of war crimes in the early 2010s by Croatian courts.

After the war and transition from socialism to capitalism, Sisak experienced rapid deindustrialization and depopulation. Many formerly state owned industries were privatized and closed down for good, causing huge unemployment. Sisak oil refinery, however, is still open and is currently in the process of transformation into a biorefinery. Sisak is in an extremely difficult position — the former industries had horrible environmental impact, but they provided the much needed jobs. Now most of the jobs are gone, but the pollution is still present.

We completed our ride on the shore of the Kupa river by the Old bridge and got some pizza. The riverfront was full of families having coffee, walking or biking. Sisak is going through some difficult times right now, but we could see that it’s a resilient town with a lot to offer. It persevered for over 2000 years, and it surely will make it through the current crisis. We had our lunch next to a group of teenage football players, getting pizza with their coach after the Sunday game. They were loud, the way teenage boys always are, but we enjoyed their company. It had been quite a lonely ride, and it was good to be in a crowded place!

On the ride back, we didn’t follow the Sava all the way, rather opting to pass through some of the bigger villages further West of the river. It was still warm, but the sun was getting low quickly since it was late October after all. We rode faster, trying to make up some time and get back to the city where our family waited for us at the “finish line” next to Maksimir park.

We passed the landmarks that were familiar to us by now, and luckily the roads remained almost as empty as they were in the morning. Eventually, we hit the A3 motorway and found the secret overpass — as soon as we crossed it we were back in the city of Zagreb! The last 10km were the hardest, and we were thankful for the protected bike route on the now busy Radnička corridor.

We made it back just in time! My parents and my sister welcomed us back at the entrance to Maksimir park. We covered a total 132km of distance, 28km shy of 100 miles, but it didn’t matter. We rode to raise funds for Multiple Sclerosis research, and together with Josh and Mechanical Gardens we raised over $3000! My dad has had the MS diagnosis for almost 10 years now, and living with this illness hasn’t been easy. It’s unpredictable, and there is so much we still don’t know about it — research is critical in finding the cure, but funds raised by Bike MS are also used to provide services to those living with MS — providing financial assistance, medical equipment, emotional care and much more. 

I hope you’ve enjoyed this ride report. If you’re able to, I encourage you to donate to the National MS Society and Make the Road NY. If you’re interested in doing a bike ride in Croatia (once it becomes possible to travel internationally again!) and if have any questions, please reach out to me over email (vladimirvince at gmail dot com) or twitter (@mejs), I would be so very happy to help you plan your trip!

The post Biking for MS in Croatia: 130km from Zagreb to Sisak appeared first on Vlado Vince.

Coney Island Polar Bear Plunge, January 1 2020

The post Polar Bear Plunge 2020 at Coney Island appeared first on Vlado Vince.

On Tuesday evening, July 30th, NYC DOT presented the plan for safety improvements and dedicated bus lanes on Church Avenue to the members of Brooklyn Community Board 14 and the interested public. Church Avenue has been in the news recently due to two deadly crashes, and it’s known to be perpetually blocked due to traffic and double parked cars. The event was hosted at Beth Shalom v’Emeth Reform Temple on Church Avenue, a fact that became a significant factor in the Q&A discussion that ensued following the presentation.

The event was organized by Council Member Mathieu Eugene (District 40). The Church Ave Transit & Traffic Improvements presentation (initially presented to CB14 Transportation Committee on May 30th) was presented by two NYC DOT staffers and Brooklyn Borough Commissioner Keith Bray. Two officers from 70th Precinct were also in attendance. Council Member Eugene explained that he called this meeting because residents had concerns about losing parking if the DOT moves ahead with installing a dedicated bus lane. After his introduction, the DOT staff presented essential information about transportation conditions on Church Avenue and the plan for transit and traffic improvements:

• The B35 is the 3rd busiest route in Brooklyn, with nearly 30,000 rides per day. 
• Bus speeds at peak times average only 4.25 mph
• Bus riders make up 72% of travelers on Church Ave during the AM peak
• Only 7% of people get to Church Ave using cars
• 72% of illegally parked cars are personal cars
• 73% of drivers parking illegally do not attempt to find a legal metered space

DOT’s plan consists of turning curbside parking between Ocean Parkway and 16th Avenue into dedicated bus lanes between 7am – 7pm, Monday through Saturday. The proposal includes additional loading zones and metered parking on side streets. The section between 16th Avenue and Flatbush, referred to by DOT as a “typical commercial core”, would have some metered parking turned into loading zones, while remaining metered parking would change from 1 hour to 30 minutes. The goal of these changes is to allow faster deliveries to businesses and to prevent double parking which is rampant on Church Avenue.

Before the DOT could finish the presentation, a few audience members started loudly complaining and booing at the mention of lost parking. The two DOT staffers handled the hecklers with respect and restraint, moving into the Q&A session moderated by Commissioner Bray, who read the questions submitted by the attendees. The questions were almost exclusively focused on parking, and the few loudest attendees tried to provoke and insult the DOT staff. Particularly unpleasant was a bad faith question about how many people with disabilities were included in DOT’s questionnaire (per DOT personal information was not collected but accessibility organizations were consulted). The attendee would not let the DOT staff answer the question and interrupted them multiple times as they tried to speak. The same attendee later went on a rant about cyclists (the DOT’s proposal doesn’t include bike lanes or any explicit bike safety improvements), complaining that the hardworking men and women of the 70th precinct waste time ticketing double parked cars because cyclists break laws (the logic of the argument was unclear). While the DOT made the best effort to pacify and appease the rowdy audience, the Q&A was getting out of control until Council Member Eugene, who had previously left the event, suddenly showed up. He proceeded to appeal to the audience concerned about parking, and promised to speak to City Council and Commissioner Trotenberg on their behalf. Most of the audience was not convinced.

The second attempt at reading audience questions was derailed after a question regarding parking for the synagogue congregants. Rabbi Heidi Hoover, who hosted the meeting, said that the congregation was not included in the DOT questionnaire that formed the basis of the data included in the presentation. Commissioner Bray answered that the DOT received the congregation’s letter and included their concerns in the proposal. A few members of the audience accused the DOT of discriminating against the congregation. There were calls for religious exemptions for metered parking. The final 15 minutes of the meeting dissolved into a mostly one-sided shouting match accusing the DOT of taking away parking and ignoring the needs of the community. There were a few members of the audience who tried to counter the  accusations aimed at the DOT, repeating that statistically only 7% of people on Church Avenue use cars, but they were mostly drowned out. Eventually, the group concerned about parking started chanting in unison “Is this a done deal?”, to which Commissioner Bray answered “We are implementing this plan”. A loud clap and a few cheers could be heard from parts of the audience. Soon afterwards the meeting ended, and Council Member Eugene grabbed the microphone and tried to speak to the audience about what he is going to do about the parking situation. Almost no-one stayed to listen.

Despite the almost absolute dominance of loud and aggressive calls to scrap the plan and keep the status quo, the unequivocal commitment by the DOT to move ahead with the plan suggests that Church Avenue will be getting the much needed transit and safety improvements. Hopefully the other dangerous thoroughfares in the neighborhood like Coney Island Avenue will be getting some attention from the DOT soon.

The post NYC DOT Church Avenue Transit & Traffic Improvements presentation appeared first on Vlado Vince.

The post Under Manhattan Bridge, May 2019 appeared first on Vlado Vince.

The post Taking a break at 10th Ave, New York, May 2019 appeared first on Vlado Vince.

The post Second Avenue protected bike lane, New York. May 2019 appeared first on Vlado Vince.

The post Coney Island Mermaid Parade, June 2018 appeared first on Vlado Vince.

The post Sunset Park, Brooklyn, October 2017 appeared first on Vlado Vince.

The post Southern Vermont, taken from Amtrak’s Vermonter train in October 2017 appeared first on Vlado Vince.

The post Near Hartford, VT. October 2017 appeared first on Vlado Vince.

Last night I made my first color prints at Gowanus Darkroom. I found this community studio space last month when I needed a darkroom to develop that efke IR film I had. They also do classes, and since I’ve never done any color film printing, I signed up for their color crash course. It was such an odd experience after years of doing B/W printing. Working in total darkness with a number of other people, yelling what you’re doing so you don’t run into someone (which of course you do). It was weird working with filters and having to wait for prints to finish processing before seeing any results, but I have to admit I love the cleanliness of the machine-aided process.

The post Gowanus canal, Brooklyn, October 2017 appeared first on Vlado Vince.

Shot on expired efke IR820 film.

Usually I post descriptions of the places I photograph. Today, however, the location is not particularly important nor interesting. These photos were shot on a roll of expired Efke IR820 film. Efke was a brand of film manufactured by Fotokemika in Samobor, a small town next to Zagreb.

Efke was synonymous with photography in Yugoslavia, and Fotokemika produced Efke film for over 60 years, using the same formula which eventually gave it a very recognizable, “retro” look. When I first started shooting film I almost exclusively shot on Efke rolls. I also learned to process film and develop prints using Efke paper and chemicals – they were the cheapest and most accessible (thanks to my high school professor who gave me the school’s leftover chemical stock!). I did not particularly consider how amazing it was that photographic film was still being made so close to me, nor how significant it was in terms of our culture, history and industrial heritage.

Fotokemika ceased all production in 2012. An attempt to save some of the equipment and remains of the old factory failed; everything that was left ended up in the dump. This summer I was looking through my old things in Zagreb and I came across an expired roll of Efke IR820 infrared film. I only shot infrared once before and I found it pretty cumbersome, but this was an unexpected opportunity to shoot a roll of Efke film one last time. These photos were shot around Brooklyn during September 2017 using an IR filter. They are somewhat blurry and messy, since I did not use a tripod. However, I hoped to gesture towards a past experience that is both personal and collective, a piece of material history that is now gone.

The post Brooklyn, September 2017 appeared first on Vlado Vince.

Baltimore, MD, September 2017

The post Baltimore, MD, September 2017 appeared first on Vlado Vince.

Zagreb, Croatia, August 2017

The post Zagreb, Croatia, August 2017 appeared first on Vlado Vince.

Risnjak National Park, Croatia, August 2017

The post Risnjak National Park, Croatia, August 2017 appeared first on Vlado Vince.

Kupa river spring, Risnjak National Park, Croatia, August 2017

The post Kupa river spring, Risnjak National Park, Croatia, August 2017 appeared first on Vlado Vince.

New York Penn Station, July 2017

The post New York Penn Station, July 2017 appeared first on Vlado Vince.

Atlantic/Nostrand, Brooklyn, July 2017

The post Atlantic/Nostrand, Brooklyn, July 2017 appeared first on Vlado Vince.

Midtown Manhattan, June 2017

The post Midtown Manhattan, June 2017 appeared first on Vlado Vince.

South Slope, Brooklyn, June 2017

The post South Slope, Brooklyn, June 2017 appeared first on Vlado Vince.

Tax March NYC, April 15th.

See more at https://medium.com/@vladimirvince/tax-march-nyc-april-15-5319970f0145

The post Tax March NYC, April 15th.  appeared first on Vlado Vince.

Midtown skyline, April 2017

The post Midtown skyline, April 2017 appeared first on Vlado Vince.

Rally and vigil for Timothy Caughman, Union Square, Manhattan, March 2017

The post Rally and vigil for Timothy Caughman, Union Square, Manhattan, March 2017 appeared first on Vlado Vince.

We Will End the Refugee & Muslim Ban March & Rally at Battery Park, January 2017

More photos here: https://medium.com/@vladimirvince/we-will-end-the-refugee-muslim-ban-march-rally-at-battery-park-402b43af5572

The post We Will End the Refugee & Muslim Ban March & Rally at Battery Park, January 2017 appeared first on Vlado Vince.

Second Avenue Subway, Manhattan, January 2017

The post Second Avenue Subway, Manhattan, January 2017 appeared first on Vlado Vince.

Jamaica Bay, December 2016

The post Jamaica Bay, December 2016 appeared first on Vlado Vince.

The 1 train, the Bronx, September 2016

The post The 1 train, the Bronx, September 2016 appeared first on Vlado Vince.

The post Park Maksimir, Zagreb, December 2016 appeared first on Vlado Vince.

Zagreb, December 2016

The post Zagreb, December 2016 appeared first on Vlado Vince.

Gundulićeva street, Zagreb, December 2016

The post Gundulićeva street, Zagreb, December 2016 appeared first on Vlado Vince.

Rockaway Inlet, Brooklyn, December 2016

The post Rockaway Inlet, Brooklyn, December 2016 appeared first on Vlado Vince.

Bay Ridge Promenade, Brooklyn, November 2016

The post Bay Ridge Promenade, Brooklyn, November 2016 appeared first on Vlado Vince.

Marine Park, Brooklyn, November 2016

The post Marine Park, Brooklyn, November 2016 appeared first on Vlado Vince.

Floyd Bennett Field, Brooklyn, November 2016

The post Floyd Bennett Field, Brooklyn, November 2016 appeared first on Vlado Vince.

BMT Brighton Line / BMT Franklin Avenue Line, Brooklyn, November 2016

The post BMT Brighton Line, November 2016 appeared first on Vlado Vince.

Gowanus Canal, Brooklyn, October 2016

The post Gowanus Canal, Brooklyn, October 2016 appeared first on Vlado Vince.

Coney Island, September 2016

The post Coney Island, September 2016 appeared first on Vlado Vince.

Verrazano Bridge, Brooklyn, April 2016

The post Verrazano Bridge, Brooklyn, April 2016 appeared first on Vlado Vince.

Zagreb Glavni kolodvor, July 2016

The post Zagreb Glavni kolodvor, July 2016 appeared first on Vlado Vince.

Deák Ferenc tér Metro station, Budapest, July 2016

The post Deák Ferenc tér Metro station, Budapest, July 2016 appeared first on Vlado Vince.

Milano, Italy, July 2016

The post Milano, Italy, July 2016 appeared first on Vlado Vince.

Downtown Manhattan, May 2016

The post Downtown Manhattan, May 2016 appeared first on Vlado Vince.

Coney Island, June 2016

The post Coney Island, June 2016 appeared first on Vlado Vince.

Brooklyn, June 2016

The post Brooklyn, June 2016 appeared first on Vlado Vince.

Lady Bird Lake under I-35 Bridge, Austin, TX, May 2016

The post Lady Bird Lake under I-35 Bridge, Austin, TX, May 2016 appeared first on Vlado Vince.

Federal Plaza, Manhattan, May 2016

The post Federal Plaza, Manhattan, May 2016 appeared first on Vlado Vince.

Bryant Park, May 2016

The post Bryant Park, May 2016 appeared first on Vlado Vince.

Coney Island, February 2016

The post Coney Island, February 2016 appeared first on Vlado Vince.

Coney Island Rapid Transit Car Overhaul Shop, Brooklyn, May 2016

The post Coney Island Rapid Transit Car Overhaul Shop, Brooklyn, May 2016 appeared first on Vlado Vince.

Flatbush, Brooklyn, May 2016

The post Flatbush, Brooklyn, May 2016 appeared first on Vlado Vince.

West Side Line tracks, Manhattan, May 2016

The post West Side Line tracks, Manhattan, May 2016 appeared first on Vlado Vince.

George Washington Bridge, Manhattan, May 2016

The post George Washington Bridge, Manhattan, May 2016 appeared first on Vlado Vince.

Verrazano Bridge, Brooklyn. April 2016.

The post Verrazano Bridge, Brooklyn. April 2016. appeared first on Vlado Vince.

Downtown Manhattan, taken from the Wonder Wheel at Coney Island. March 2016.

The post Downtown Manhattan, taken from the Wonder Wheel at Coney Island. March 2016. appeared first on Vlado Vince.

Washington Cemetery, Brooklyn. Ferbruary 2016.

The post Washington Cemetery, Brooklyn. Ferbruary 2016. appeared first on Vlado Vince.

Park Slope, Brooklyn. Ferbruary 2016.

The post Park Slope, Brooklyn. Ferbruary 2016. appeared first on Vlado Vince.

Coney Island, November 2015.

The post Coney Island, November 2015. appeared first on Vlado Vince.

Edinburgh, September 2015.

The post Edinburgh, September 2015. appeared first on Vlado Vince.

2nd Avenue Station, Manhattan. January 2016.

The post 2nd Avenue Station, Manhattan. January 2016. appeared first on Vlado Vince.

Prospect Park, Brooklyn, January 2016

The post Prospect Park, Brooklyn, January 2016 appeared first on Vlado Vince.

Texas Capitol Building, Austin, TX. December 2015.

The post Texas Capitol Building, Austin, TX. December 2015. appeared first on Vlado Vince.

East Side, Manhattan. November 2015.

The post East Side, Manhattan. November 2015. appeared first on Vlado Vince.

Coney Island, Brooklyn. November 2015.

The post Coney Island, Brooklyn. November 2015. appeared first on Vlado Vince.

Brighton Beach, Brooklyn. November 2015.

The post Brighton Beach, Brooklyn. November 2015. appeared first on Vlado Vince.

Rovinj, Croatia. 2015

The post Rovinj, Croatia. 2015 appeared first on Vlado Vince.

Hacklebarney State Park, NJ. 2015

The post Hacklebarney State Park, NJ. 2015 appeared first on Vlado Vince.

Inverness, Scotland. 2015

The post Inverness, Scotland. 2015 appeared first on Vlado Vince.

Scaniport, Higlands. Near Loch Ness. 2015

The post Scaniport, Higlands. Near Loch Ness. 2015 appeared first on Vlado Vince.

Corona Yard, Queens. 2015

The post Corona Yard, Queens. 2015 appeared first on Vlado Vince.

Monument to the Revolution of the people of Moslavina. Podgarić, Croatia. 2015

The post Monument to the Revolution of the people of Moslavina appeared first on Vlado Vince.

Pula, Croatia. 2015

The post Pula, Croatia. 2015 appeared first on Vlado Vince.

Oslo. 2015

The post Oslo. 2015 appeared first on Vlado Vince.

New River Trail, Virginia. 2015

The post New River Trail, Virginia. 2015 appeared first on Vlado Vince.

Fort Tilden, NY. 2015

The post Fort Tilden, NY. 2015 appeared first on Vlado Vince.

Bay Ridge Branch, Brooklyn. 2015

The post Bay Ridge Branch, Brooklyn. 2015 appeared first on Vlado Vince.

Central Park. 2015

The post Central Park. 2015 appeared first on Vlado Vince.

Greenpoint, Brooklyn. 2015

The post Greenpoint, Brooklyn. 2015 appeared first on Vlado Vince.

Newton Creek, Brooklyn. 2015

The post Newton Creek, Brooklyn. 2015 appeared first on Vlado Vince.

Washington Square Park, New York. 2015

The post Washington Square Park, New York. 2015 appeared first on Vlado Vince.

Brighton Beach, Brooklyn. 2015

The post Brighton Beach, Brooklyn. 2015 appeared first on Vlado Vince.

Brooklyn Bridge Park, 2015

The post Brooklyn Bridge Park, 2015 appeared first on Vlado Vince.

East River, 2014

The post East River, 2014 appeared first on Vlado Vince.

Hudson River, 2014

The post Hudson River, 2014 appeared first on Vlado Vince.

Fourth of July Parade, Montclair, NJ. 2014

The post Fourth of July Parade, Montclair, NJ. 2014 appeared first on Vlado Vince.

Hackensack River with Pulaski Skyway and Kearny Power Plant in the background, NJ. 2014

The post Hackensack River with Pulaski Skyway and Kearny Power Plant in the background, NJ. 2014 appeared first on Vlado Vince.

Execution Rocks Light, Long Island Sound, NY. 2014

The post Execution Rocks Light, Long Island Sound, NY. 2014 appeared first on Vlado Vince.

Route 66, TX. 2014

The post Route 66, TX. 2014 appeared first on Vlado Vince.

Cloudy General Store. Open sometimes. Cloudy, OK. 2014

The post Cloudy General Store. Open sometimes. Cloudy, OK. 2014 appeared first on Vlado Vince.

Williamsburg, 2014

The post Williamsburg, 2014 appeared first on Vlado Vince.

Portland, OR. 2014

The post Portland, OR. 2014 appeared first on Vlado Vince.

Budapest, 2014

The post Budapest, 2014 appeared first on Vlado Vince.

Broadway/Sixth ave, New York. 2014

The post Broadway/Sixth ave, New York. 2014 appeared first on Vlado Vince.

East River, 2014

The post East River, 2014 appeared first on Vlado Vince.

Zagreb, 2014

The post Zagreb, 2014 appeared first on Vlado Vince.

Lower East Side, Manhattan. 2014

The post Lower East Side, Manhattan. 2014 appeared first on Vlado Vince.

Zagreb, 2014

The post Zagreb, 2014 appeared first on Vlado Vince.

Mazatlán, Mexico. 2013

The post Mazatlán, Mexico. 2013 appeared first on Vlado Vince.

Chicago, 2013

The post Chicago, 2013 appeared first on Vlado Vince.

neka glupa koza

The post neka glupa koza appeared first on Vlado Vince.

Vera

The post Vera appeared first on Vlado Vince.

Dobrinj, 2013

The post Dobrinj, 2013 appeared first on Vlado Vince.

Amsterdam, 2013

The post Amsterdam, 2013 appeared first on Vlado Vince.

Uvala voz, Krk, Croatia. 2013

The post Uvala voz, Krk, Croatia. 2013 appeared first on Vlado Vince.

Catalina, 2013

The post Catalina, 2013 appeared first on Vlado Vince.

Sarajevo, 2013

The post Sarajevo, 2013 appeared first on Vlado Vince.

Visoko, BiH. 2013

The post Visoko, BiH. 2013 appeared first on Vlado Vince.

Zagreb, 2007

The post Zagreb, 2007 appeared first on Vlado Vince.

Zavrtnica, Croatia. 2013

The post Zavrtnica, Croatia. 2013 appeared first on Vlado Vince.

Middle of nowhere, TX. 2013

The post Middle of nowhere, TX. 2013 appeared first on Vlado Vince.

Austin, TX. 2013

The post Austin, TX. 2013 appeared first on Vlado Vince.

Kanader, Canadair. 2012

The post Kanader, <em>Canadair. </em>2012 appeared first on Vlado Vince.

Zagreb, 2013

The post Zagreb, 2013 appeared first on Vlado Vince.

Portland, OR. 2013

Ross Island Bridge, I-5 Marquam Bridge and Portland-Milwaukie Light Rail Bridge under construction

The post Portland, OR. 2013 appeared first on Vlado Vince.

Zagreb. 2013

The post Zagreb. 2013 appeared first on Vlado Vince.

Ljeto, ne spominji. 2009.

Negdje na Jadranu, normalno.

The post Ljeto, ne spominji. 2009. appeared first on Vlado Vince.

Pietro, 2013

The post Pietro, 2013 appeared first on Vlado Vince.

Drugi kut. Beograd, 2013

The post <em>Drugi kut. </em>Beograd, 2013 appeared first on Vlado Vince.

Beograd. 2009

The post Beograd. 2009 appeared first on Vlado Vince.

Tillamook Bay, Oregon. 2011

The post Tillamook Bay, Oregon. 2011 appeared first on Vlado Vince.

Sarajevo, kristalna sećanja. 2011 (fotografirano pri povratku, negdje u zoru pred Zagrebom)

The post Sarajevo, kristalna sećanja. 2011 appeared first on Vlado Vince.

Vancouver, BC. 2012

The post Vancouver, BC. 2012 appeared first on Vlado Vince.

Slavonija/Vojvodina? Srbija/Hrvatska/Jugoslavija? 2013

The post Slavonija/Vojvodina? Srbija/Hrvatska/Jugoslavija? 2013 appeared first on Vlado Vince.

Zrinjevac, Zagreb. 2013

The post Zrinjevac, Zagreb. 2013 appeared first on Vlado Vince.

Maksimir, Zagreb. 2013

The post Maksimir, Zagreb. 2013 appeared first on Vlado Vince.

Maksimir, Zagreb. 2013

The post Maksimir, Zagreb. 2013 appeared first on Vlado Vince.

Rome? 2009

The post Rome? 2009 appeared first on Vlado Vince.

Adam, Oregon Coast. 2012

The post Adam, Oregon Coast. 2012 appeared first on Vlado Vince.

Revolution/Šta je ostalo. 2011

The post <em>Revolution</em>/Šta je ostalo. 2011 appeared first on Vlado Vince.

Killing Moon/Mesec žut. 2012.

The post Killing Moon/Mesec žut. 2012. appeared first on Vlado Vince.

Zach, Skewed Concoctions. Portland, Oregon. 2012

The post Zach<em>, Skewed Concoctions</em>. Portland, Oregon. 2012 appeared first on Vlado Vince.

Oregon coast. 2012

The post Oregon coast. 2012 appeared first on Vlado Vince.

Sestrica Vela (Big Sister Island), Croatia. 2012. One of the many lighthouses on the Adriatic coast available for rent.

The post Sestrica Vela (Big Sister Island), Croatia. 2012. appeared first on Vlado Vince.

Oregon Route 35. 2012

The post Oregon Route 35. 2012 appeared first on Vlado Vince.

Eastmoreland. Portland, Oregon. 2012

The post Eastmoreland. Portland, Oregon. 2012 appeared first on Vlado Vince.

I-84. Portland, Oregon. 2012

The post I-84. Portland, Oregon. 2012 appeared first on Vlado Vince.

ferry. Cres, Croatia. 2012

The post ferry. Cres, Croatia. 2012 appeared first on Vlado Vince.

Brodovi, 2007

The post Brodovi, 2007 appeared first on Vlado Vince.

Wooden Skyscraper, Zagreb, Croatia. 2012

The post Wooden Skyscraper, Zagreb, Croatia. 2012 appeared first on Vlado Vince.

Portland, Oregon. 2012

The post Portland, Oregon. 2012 appeared first on Vlado Vince.

Rome. 2009

The post Rome. 2009 appeared first on Vlado Vince.