Make Hack Learn

Visiting Metrix

2014-04-30T12:00:00-07:00

Last week I was up in my old home of Seattle for a few days to see family and friends (my daughter just turned 17).

While there, I dropped by Metrix Create:Space on and off as I was staying a 10 minute walk away and the owner, Matt, is a friend of mine. I talk to Matt on IRC fairly often and have been monitoring 3D printing developments at Metrix through him and watching their flickr feed. Since I had the time, I came to the weekly 3D printing night to meet folks working on projects at Metrix.

Johann Rocholl has been the primary moving force in the current interest in Delta printers within the reprap community. He has a tumblr with many pictures of his work and maintains a github repo for development. Terence Tam also does a lot of work at Metrix with his OpenBeam aluminum extrusion, which is a 15 mm profie extrusion that is open source and used by a lot of folks for building 3D printers. There is also Mattew Wilson, who has developed the Brainwave all in one printer control board (which is very nice), working out of Metrix.

They have quite a critical mass of people doing fun things with 3D printers and open source.

You can see the "OpenBeam Kossel Pro" that Terence and Matt have been developing:

Johann was also working on his "Mini Kossel" (which also uses OpenBeam), which is his attempt to build a portable Delta printer that can travel easily and maybe even be battery powered:

From talking to folks, I expect that there will be announcements at Maker Faire in Redwood City in a few weeks and Delta printers, among others, will make a heavy appearance there.

It was nice to see folks actively working on printers and advancing the technology associated it.

I haven't posted much about my printer projects in the last few months as I've been making incremental progress. My Rostock Mini is largely stalled out due to lack of desire to deal with some design issues on it (though I may pick it up again soon) but I have a Hadron Ordbot that is completed except for mounting the heated build platform and splicing some wires. I've tested all of the electronics and motion but haven't printed with it yet. The Foldarap has been waiting for the last six weeks or so for me to mount its printing bed and carriage but I hope to have it on the Ace Monster Toys table at Maker Faire in three weeks.

current state of my Foldarap

my Hadron Ordbot

One thing that I've been doing is co-hosting an ongoing 3D printing meetup every other Wednesday night at Ace Monster Toys. People have been showing up to work on printers, discuss issues around them, and generally show off their work. I'm actively working on creating a bill of materials to self-source parts to do a workshop series where 10 people (hopefully) build a Printrbot Jr. clone at AMT. The real issue there is trying to get the cost of the materials as near a $250 price point as possible (and it may not be possible to get below $300 really). Electronics from Printrbot are $129 retail, the hotend is $59, and, realistically, we need four stepper motors for roughly $15 each plus all the screws, rods, etc. It is the combination of the electronics, hotend, and motors that is kind of hard to move without just completely replacing them with someone else's parts.

Atom Bomb at AMT is actively discussing the development of a new "all in one" 3D controller board with built in stepper controllers that do 1/16 motion. We're hoping to publish a specification and have PCBs made for this, even if we don't use this for the 3D printing workshop. The end-goal is to have a solid board with a total cost for the bill of materials around $40, which is half or a third of the common cost for RAMPS and other 3D controllers. I expect we'll have more news on this in a few weeks or a month.

I'm hoping to have more announcements in the future but all of this means that I'll probably be assembling a Jr. (or a variant) soon because I'm going to need to know it backwards and forwards to teach a class on it and I may wind up making a bit of a variant later.

Foldarap Adventures

2013-01-19T21:30:00-08:00

When last I wrote, I said that I was building a Rostock Mini 3D printer. That is still true but I hit a few snags. The creator, who did a lot of cool work making a parameterized design for it, never quite published his extruder design. That means that I've been trying to figure out how I was going to get the business end of the printer extruding plastic. I've been looking at v3 of the Airtripper Bowden Extruder but had to order a bunch of parts for it. (As an aside, while I do everything in metric to maintain compatibility with the worldwide making community, it is a real pain to get metric screws of all sizes in any number so I have to order them and wait wait wait…) This combined with needing to replace the carbon pieces that I was using for arms meant that I was a little stalled out on the Rostock Mini.

Foldarap by Emmanuel Gilloz

So, like any dumb hacker, I decided to work on another printer instead. I guess I've caught the reprap bug. I decided to make a FoldaRap. Emmanuel Gilloz, a French hacker, came up with his own RepRap design that uses 20 mm extruded aluminum (which is cheap) that can be folded up and put into a padded case. This makes it a both affordable design but one that can easily be taken to hackerspaces, conferences, or other events. Since I've had to lug my Up! Plus to Ace Monster Toys three or four times in the foot well of my car, this is a real plus.

Foldarap in Suitcase by Emmanuel Gilloz

Building this, the most difficult part has been sourcing the materials. Emmanuel is in France and they have a different convenient supply chain in Europe. He published his build of materials but I had to do my best to make some adjustments (and find yet more metric screws). The two biggest issues were the source of the aluminum extrusion and the fact that he used some RepRap Huxley components, including a Huxley hotend.

With the Huxley components, that meant that he used a 140 mm square aluminum heating bed. Since Prusa Mendels are the most common RepRaps that I see here in the US, I either had to order overseas for the bed or cut my own. I actually found an individual making them, removing the necessicity to cut my own plate and then tap it for screw holes for mounting.

Most American builders are using hotends from Makergear or J-Head hotends (which are out of stock for my size plastic except for Chinese copies). I didn't feel inclined to deal with yet another custom extruder system (see Rostock Mini issues at beginning of post) so I wound up getting the Huxley extruder that Emmanuel recommended, minus a few unnecessary parts.

The extrusion of choice here in the US is MiSumi 2020. It costs $3 for a 300 mm long piece, which means for about $30 plus cheap shipping (in California even), I can have regular, solid aluminum pieces. As a bonus, Misumi will cut to length in half millimeter increments so the pieces arrive ready to go.

All of my materials arrived about a week ago except for my Huxley hotend and my aluminum (with the exception of my control electronics). I used my existing printer to start making pieces per the FoldaRap design and immediately hit a snag: my plastic wouldn't fit on my extrusion. (No, that isn't a euphemism.) I asked Emmanuel about it on the RapRap forums and he was very responsive to questions. Initially, it seemed that my Up! wasn't printing with enough precision to fit the gap between my extrusion but, pulling out my calipers and then looking at Emmanuel's designs, I found that his design had a 1.5 mm gap where my extrusion was (roughly) 2 mm thick:

not 2 mm!

Luckily for me, open source wins! Emmanuel has the source for the FoldaRap up on github. Like many hackers, he does all of his 3D design in Sketchup (because it is free and easy to use, I assume). I've never done much with it but I've seen it used at AMT quite a bit by a few of our members. A quick download of it and the github source and I had parts to stare at. I then spent an hour givining myself a crash course in how to alter metric parts in Sketchup, expanding the gap in the above picture by half a millimeter on four pieces that needed to fit with a plastic 'T' connector. Exporting it out to STL format, I started a test print and went to dinner with my wife. On returning home, I tried it out and it worked!

If Emmanuel wants it, I'll submit a patch to Github with the Misumi specific changes to the four files. I made a separate sketchup file with just those four pieces in it. That might be my first submitted patch on github if I do so.

I'm now printing out the other three pieces (two of them are three hour prints each) so I can begin the actual process of assembling the FoldaRap over this three day weekend. Given the missing electronics and general slackitude, I don't expect that I'll finish it this weekend (hmm…I'm missing my power supply as well, come to think of it) but I do expect progress. I'll then get back to my pesky Rostock Mini. I expect that the FoldaRap build will go quite a bit more quickly as Emmanuel's build documentation is quite thorough with both pictures and video of the various stages.

Building a Rostock Mini

2012-12-31T16:00:00-08:00

I haven't posted in a while here, largely because I haven't finished any interesting projects lately. The one thing that I've been building recently is a delta printer. This is a slightly different design than the standard for a reprap printer.

Johann, up in Seattle, has been playing with these during the last eight months or so. He published an initial design on Thingiverse, which led to a lot of people getting excited and working on the project. (In fact, there is a very active Google Group for it, right now.)

You can see his original version doing a print below:

For those familiar with reprap printers, this is a very different model for moving the printing head.

Brian Evans in Colorado came up with a variant, called the Rostock Mini (or here. Originally, it was meant to be a more desktop sized version of the Rostock, as they can get quite large. In the process of making it, Brian created an OpenSCAD version of it that is parameterized, so you can create an arbitrarily sized version. It turns out, it isn't really that desktop sized but it more convenient.

Right now, I have an Up! Plus printer, which is a closed source Chinese printer. It is pretty reliable but you can't do fairly standard things, within the open source reprap community, like tweaking the slicer settings or temperatures. While I've enjoyed using it, I've found it frustrating enough that I wanted to build a new reprap printer. Seeing Brian's design, I decided to start on a Rostock Mini a couple of months ago. It took a while for me to get parts printed, then I had to track down all of the various metric screws, bolts, etc., as well as electronics for it and steppers. If there is a complaint that I have about making reprap printers, it is that by the time you source all of the parts from various places, it winds up being at least 30% more expensive than you expect and you often sit around waiting for a single piece that you're missing. In fact, at this moment, I'm still waiting to track down pieces for my Bowden extruder, needing just one or two of several screws which are sold in bags of 50 or 100 for $10.

That said, I made significant progress on putting everything together the other day, as you can see below. I printed out most of my plastic components in a colorful purple PLA plastic. I found some cast off pieces of smoking acrylic at Ace Monster Toys (they were corners left over from a larger project) and cut them on our lasercutter, along with some cork for dampening motor vibration. This weekend, I got the basic frame assembled except for the extruder platform and the three carriages where it mounts to the frame.

You can see my whole photo set for the build so far over on Flickr.

The next steps are to get the extruder platform built and the hot end mounted, then connecting belts to motors, and mounting electronics. Finally, I'll have to get the Bowden extruder mounted to it and connected.

I figure it might take me another month to get it built as I'm not in a rush but it looks pretty good so far and I'm looking forward to seeing how well it prints.

OMFG Closed Source 3D Printers

2012-09-20T11:40:00-07:00

Bre Pettis

Wow. I posted about the new Makerbot Industries Replicator 2 yesterday. I'm still, site unseen except for the videos, impressed by the design and taking things to a prosumer level but Bre and MI have apparently confirmed that the software and (it seems?) the hardware are both closed source, at least according to Josef Prusa, and people are having an Internet style freakout.

Since the first Makerbot machine, the Cupcake, and all subsequent ones, were iterations off of designs from the RepRap community. RepRap is pretty much the epitomy of an open source community in both hardware and software. Earlier versions of the Makerbot software, ReplicatorG, just wrapped up existing open source RepRap software components. There has always been some tension between Makerbot Industries as a company selling itself and its ideas to the public (and investors) and the open source roots of their work but they have handled it fairly well until now. Bre and others there have been involved publicly in various open hardware events. MI released their designs for printers to the community, making them available in repositories and on Thingiverse, their site for hosting designs. This has led to quit a few knockoffs and a few outright clones, such as the Tangibot that had a failed kickstarter and was simply a direct, Chinese manufactured, clone of the existing MI Replicator. People have wondered if MI was going to continue to be an open source company in the face of threats like Tangibot, especially with having taken VC money in the last year.

Because of this history and the reactions of many open source folks to commercialization of FOSS work, people often seem to have an ambivalent relationship to Makerbot Industries. I've been one of them, having felt burned by the way they abandoned support of existing product lines almost immediately on shipping new versions, which is not what you expect from a commercial company with customers. At my hackerspace, this has made us unwilling to rely on MI for our printers, causing us to look at and use various alternatives (and to not replace our aging Cupcakes with newer MI designs).

Today, the shitstorm has landed (more of a shitnami as I write this). Josef Prusa, of "Prusa Mendel" fame, claims that he's called MI and exchanged email with Bre Pettis and that it is confirmed that the new Replicator 2 and Makerware, their new software, are closed source. I'm not sure if this understanding is actually true or if it is that the software is closed source (it has a spiffy giant EULA) with open source components and the Replicator 2 reference designs will be open sourced once they've actually shipping in a month or two. The latter is what they did with previous designs, not making them available until after they ship. Regardless of the final picture on this, the RepRep and 3D printing community is freaking out today.

The comment thread on Josef's post is pretty epic (and his "Occupy Thingiverse" post on Thingiverse.com) and now there are a series of Google+ proclamations and responses (see 1, 2, and 3) as people start freaking out about the not-very-recently-changed Thingiverse EULA since Thingiverse is operated by MI. Josef's post and the controversy have also made Slashdot and Reddit today with people making various proposals for moving the hosting of designs off of Thingiverse.

This ought to get interesting but I do hope people calm down. Even if MI and Bre have decided to go closed source, it doesn't necessarily really affect the RepRap (or larger 3D printing community), given that 3D printing has been dominated by fairly large, closed source, corporate entities for decades. People can continue to iterate on designs and improve things. Personally, as much as I wish that it was open source, having an affordable (under $3,000) 3D printer that is reliable and usable by non-geeks will raise the bar for the larger community. It gives people a target to beat and to do so with open source. I do think people are up to the challenge and I look forward to seeing their work.

Update: Bre has posted an official reponse to the kerfuffle.

New Makerbot Printer

2012-09-19T13:40:00-07:00

Wow, today is an unexpected news day for 3D printing. Over at Ace Monster Toys, we've had a number of 3D printers, either in the space (Makerbot Cupcakes, part of a Prusa Mendel, and a Printrbot LC) but have always found them a bit of a pain in the ass, frankly. By and large, they haven't lived up to the hype, mostly because of maintenance issues which make them a bit unreliable for the kind of day in and day out printing by random folks that you expect to have in a hackerspace.

We've been talking about getting an Up printer for a while. The downside of the Up printer is that it isn't open source. It is a Chinese made printer using the ideas from the RepRap community but using closed source software. It isn't a tinkerer's device. A lot of people who have 3D printers have them as a project in and of themselves, which makes them fun to work on (I guess). If you just want to make a 3D file and print it, you're less interested in the printer as its own project and just having it work. By a number of accounts of friends at professional prototyping shops and research groups, the Up just works. It feels bad since it betrays the open source ethos.

Because of all of this, we've gone back and forth on this. I still expect that we're likely to get an Up at AMT, based on price and reliability. The wrinkle today is that Makerbot Industries, who was the first company to really comercialize RepRap hobbyist technology has finally created a printer that they consider a "prosumer" device, the Replicator 2. They announced it today in a live press event and their website has now been updated to reflect the new products.

My issues in the past with Makerbot Industries is that they tend to orphan users of their devices and have an annual or less update cycle. So, if you buy your new shiny printer kit from them, build it and are working with it, Makerbot Industries tends to end support a few months after you bought it in order to suddenly unveil their new printer. They don't do legacy support, except within the community (not their staff), which really sucks. When I owned one of their Cupcake printers, I felt really burned by this. I know, for example, that Hacker Dojo in Mountain View just ordered a Replicator printer, which was the new hotness less than six months ago at Maker Faire, just a week or so ago. Suddenly, their printer is now old news and probably won't be supported in six months.

So, this is the downside of a quickly iterating hardware startup.

The upside of quick iteration is that the state of the art seems to really improve quickly, which is the case today. The Replicator 2. It has a 0.1 mm layer height on the parts it prints out, which is damn good out of the box. I've seen other printers do this or better but not without a massive amount of tweaking and playing with them. This printer has gone from ye olde lasercut plywood, the standard for almost all, to a metal framed, factory assembled printer.

Features that they are listing are:

1.2 L x 6.1 W x 6 H in(12.75 in diagonal) buid area (which is huge)
100-micron layer resolution
Wear-resistant, oil-infused bronze bearings
Black powder-coated steel frame
Optimized for PLA (which is a nicer material in the long run)
New software (Makerware) which seems to be compiled code and optimized for use by normal humans (downside is that they are using a new file format...)

I'm actually excited by this device, which is unusual for my jaded self at this point. My main, personal, goal with a 3D printer is to design and print cool things. I'm actually quite tired of the ongoing "build and maintain a 3D printer from scratch or a kit" type projects. I just want to print stuff. This new Replicator 2, along with the current Up Plus and the upcoming Up Mini, is one of the few things out there to fulfill this goal.

The one downside is that Makerbot Industries' prices keep going up. The Cupcake was something like $900 or $1,000 as a kit and the common price point is from there to about $1,400. This new device is $2,200, basically. That isn't cheap, which means people really need to think about whether they want or need this device. It is definitely worth considering though.

Update: I see Wired magazine has a big piece up now. Make Magazine has a mini review up as well.

Making Clocks

2012-09-13T14:40:00-07:00

Today, a batch of Ardunino compatible nixie tub boards arrived from China. I heard about these a week ago on Professor Craven's blog and immediately ordered four of them for myself. For those that don't know what a Nixie Tube is, Wikipedia has the answers. As it says:

A Nixie tube is an electronic device for displaying numerals or other information. The glass tube contains a wire-mesh anode and multiple cathodes, shaped like numerals or other symbols. Applying power to one cathode surrounds it with an orange glow discharge. The tube is filled with a gas at low pressure, usually mostly neon and often a little mercury or argon, in a Penning mixture.

Basically, Nixie tubes were used before LED displays were invented to do alphanumeric displays on equipment, especially in the various militaries of the world. The numeric only ones, such as what people use for clocks, have the numbers 0 through 9 individually outlined, one on top of the other, with each one lit in turn as needed. If you've watched old 1960s science fiction movies, you've undoubtedly seen a Nixie tub display on a computer or piece of "high tech" equipment.

Professor Craven put a little video up yesterday of his tubes plugged into a board:

I have a flickr photo set of one of the boards as received.

Like Professor Craven, I plan on using them for a clock of sorts. More specifically, I plan on making a meditation timer for shits and grins. The idea is that I will laser cut a box at Ace Monster Toys to hold the hardware. I'll put an arduino and the Nixie tube boards in it, with two or more of the tubes poking out, as a display. I'll probably use two to signify minutes. I'm then thinking of adding two solenoids that go out the sides, underneath to small metal bars or rods and a few buttons. The idea is that you can set a series of times on the clock, like a sitting period and then a walking period for meditation (or a series of these as pairs) via the buttons. When each period ends, the solenoids will fire a series of times (like two or three or one, depending on what just ended or if we're completely done), which will strike the metal bars, triggering a chime.

This is similar to the "Zen Alarm Clock" that has been around forever but a bit less new agey and more…something.

This seems both appropriately hackerish or geeky and appropriately Buddhist, in some sense.

Mini Quadcopter fun at AMT

2012-09-13T12:00:00-07:00

Tim O'Brien's mini-quadcopter

Last night, we had our new semi-regular small open house at Ace Monster Toys. One of our members, Mark, had recently met another fellow, Tim O'Brien, and invited him to come by. While playing with a tiny toy (indoor) quadcopter that I picked up this week, Tim showed his scratch built mini-quadcopter. I gather from what he said (and what I see on his site), that he's been building and creating quads for couple of years now. Most of the ones that you see are much larger, roughly 18 inches or so across and actually kind of awe inspiring if you turn them on in an enclosed space (which I don't recommend). Some of us have been interested in mini or micro-quads because we'd like to be able to play with them inside of places like AMT or our homes and not just take our big flying lawnmowers to the park to fly. The fact that Tim had hacked a relatively nice one together of just that size made it rather fun to see and it was a nice coincidence that he showed up and brought it. Tim has a post on what I assume is a previous iteration of the same design, in which he built his own platform for it for a class.

I took a brief phone video of him flying his in the space (pardon my shaking cam hands) that gives a sense of the size of his quadcopter:

Since AMT has one member who regularly flies quads and something like four or five of us that have been building them but never really flown them, having someone with a few years of experience, including building from scratch, would be nice. (My big quad is fully built but completely uncalibrated for flight, for example, and Tim offered to help me calibrate it quickly.) I've asked Tim to come do a short presentation or introduction to quadcopters for some evening. We're going to figure out schedules so he can come to AMT for an hour or so some evening and talk about his work.

Notes from a Hackerspace Panel

2012-07-28T22:30:00-07:00

Drawing by Suzanne Forbes

I attended a panel today at Defcon 20 titled, "Connected Chaos: Evolving the DCG/Hackspace Communication Landscape." I'm not sure what I expected it to be but it wound up being a panel of some East Coast and Southern hackerspace folks talking about what had and hadn't worked in running a hackerspace, based on their hard won experience over the last few years. It was none of the normal West Coast folks, which made it a bit more interesting to me. Much of it jived with my own experience as a hackerspace founder so I thought that I'd share some of my loose notes.

If all of the members of the panel introduced themselves, I didn't hear it. According to Defon scheduling, they were blakdayz, anarchy angel, anch, Dave Marcus, Nick Farr. I think some of those may be pseudonyms but you never know.

Notes

Definition: "Hackerspace: A physical space where hackers gather to hack things together."

In other words, no need to get fancy and argue about what the word means.

There are six common traps for hackerspaces:

"We don't do that here"
Clubhouse/Partying
Leader-with-a-vision
Single-point-of-failure
Tasmanian Devil Trap
OMG Moneyz

Some details of what these mean:

"We don't do that here" - Hackerspaces should be working outside their comfort zone. Don't limit what people do just because people haven't done it before. Software geeks should try some electronics or making things, etc.

"Clubhouse/Partying" - This is when hanging out becomes primary over actually hacking on things. This is a really common problem. This takes it from a hackerspace to just a bunch of geeks hanging out and isn't very inviting to new folks (nor constructive).

"Leader-with-a-vision" - A strong founder with a vision starts the space and it becomes trapped by this leader's vision. New people and new ideas are essential over time as things adjust and change. The founder should eventually step aside completely to just become another member.

"Single-point-of-failure" - Key leaders or admins are points of failure. If there are no backups or replacements for folks then the group can be limited by the loss of a person to the group. You should be finding and training your replacement if you are in charge of something. This is a variation of "Dave got hit by a bus" as a problem.

"Tasmanian Devil Trap" - If you're not learning, teaching, hacking, or figuring out a problem, you shouldn't be there. People should not spend all of their time talking or arguing about a person or thing, circling around and around. That person or thing should go and let people hack. Otherwise known as "Drama."

"OMG Moneyz" - This is a hyperfocus on ways to make money for the space, to pay bills or otherwise. It can become a perennial distraciton. The solution to money issues is to attract more people and your money problems go away (because new people support the space, join, pay dues, etc.). Come up with classes, activities, and so forth to get people to show up and want to do things as a community.

Second commentor (whomever it was) said "Hackers are inherently lazy. You have to motivate people to want to do stuff."

Dave said "Consistency." You need to be consistent in the space. Have regular classes, a consistent schedule of classes or events. If you don't have this, it becomes a clubhouse very quickly. He saw this happen at his hackerspace. He also says "Location, location, location." Pick the right place for your space. This cannot be over emphasized. He also says that the financial model is very important for the space.

Another commenter says that there needs to be space to work. Don't fill up the space with junk. Don't become the final drop off point for other people's crap. He related coming to the space one day to find racks and servers piled up to the ceiling, all useless. WTF are you supposed to do with useless junk? People want to bring stuff but it can't sit around hoping for a project.

If you do have a big pile of junk, take it apart. You might learn something when you do. Pull out the useful parts and pieces, keep those, and get rid of the rest.

Dave also says that he thinks a good hackerspace starts more projects than they finish and that's ok. People should be encouraged to be creative as long as the debris doesn't take over the space. Better to create than to not do anything at all.

"You want to teach something? Get up and teach it."

"What people do in the space sets the tone for the space and its future."

There was advice, in response to an audience question on dealing with the press and local community - a few things that work well:

Write Press Releases - reporters are lazy/overworked. Write press releases for them and they'll often run with them. Make friends with the local journalists and editors. They are in your community.
Get excited about projects at other hackerspaces - if you see cool things, reach out, tweet about them, blog about them, etc. Make connections with these spaces and support them and they will support you.
Reach out to people at cons (like Defcon, I guess, meh.).
Consistency in media/Internet publication: lots of tweeting, blogging, etc. on a regular basis. Don't go dark. Put information out there and make your presence known.
Get other allied groups to meet in your space - 2600 groups, Linux User Groups, etc. Give them space to meet and bring people in. They may join your community and are natural interest groups.
Participate in your local community, not just in technical ways. Help with a soup kitchen, clothing drives, things like that. Interact with non-geeks.

That's it for the notes. Maybe they'll be useful to someone!

Black Hat and Defcon Fun

2012-07-28T15:33:00-07:00

I've spent the last few days at the Black Hat and Defcon security conventions in Las Vegas. I attend these every year for my work on the security team at Mozilla. This year, somehow, I seem to have wound up as one of the responsible adults, instead of an attendee, so I helped our group of 14 or so get tickets, hotel rooms, and so forth.

The talks this year have been fairly hit or miss but that is true every year. For all of the chatter about it, Black Hat is only a two day conference, so the amount of things presented there is often quite limited. Defcon, for all of its pretensions of being the "hacker" convention as a counter to Black Hat being the "professional" one, contains much of the same people, is three days long instead of two, and often has more interesting talks.

It is clear that phone hacking is still a big area of focus for people. As in recent years, there have been a lot of talks that look at hacking Android and iOS devices in various ways. Anyone who follows security news will have heard that Apple actually officially sent a presenter to speak on iOS for the first time ever (though I wasn't in that talk) and there have been a number of other presentations on iOS. The focus at Defcon so far has been more on hacking Android from what I've seen. Various people looking at malware on Android and ways to pwn the system. I attended a somewhat interesting talk on folks aborting the Java portion of an Android phone's system (which is where Android, itself, lives) and running a Linux distro, Inferno, on top of the stub of a system that instantiates Android. Very clunky but interesting in a geeky sort of way. I also just attended a talk on setting up a decent ad-hoc mesh network using transparent proxies running on Android systems (leaving aside the man-in-the-middle possibilities there). That one actually showed a lot of promise for networks during crises or disasters. I attended two similar talks last year and this one shows a lot more promise and robust design that those earlier presentations. You can actually see the slides and paper on Github, along with other code related to the project.

The two things that I've had the most fun with so far were not directly related to my Mozilla work in any way.

Yesterday, Chris Anderson of Wired Magazine and DIYDrones did a nice overview of the current state of the hobby drone market, what's going on, and discussed the business model and direction of 3D Robotics, the open source hardware company he founded to support what initially started as a hobby for him. He commented that economists always ask him about his business model as if it was something horribly complex or obtuse when, as he says, a 17th century fishmonger could understand it. They sell things (drone electronics and parts) for more than they cost. He said that the sweet spot for them is selling at 2.6 times the cost of the bill of materials. This leaves them enough space to sell at a margin as wholesalers to retailers and for the retailers to then do be able to do the same without trying to go crazy making things expensive. Because they have a huge DIYDrones community, they can do open source developments where they iterate quickly on improvements to the electronics (even making whole new platforms) and get them to market. The most successful and active of their community members can even get equity in 3D Robotics in return for their work on projects, making it a "win win" situation for everyone.

This morning, Cory Doctorow, of Boing Boing and writing fame, did a follow-up to his Chaos Communications Conference 28 talk on the "The Coming War on General Purpose Computing." This talk was "Beyond the War on General Purpose Computing: What's Inside the Box?" This talk focused on the after events of that war, assuming that we win it and general purpose computing doesn't go away. In it, he discusses his actual support (surprisingly) of the "Trusted Platform Module" in computing but with a twist. He focused on how it could be used to provide freedom for users by allowing them to both select and know what software (including operating systems) are running on their own machines, given people a potential for longterm control and security in the face of malware, root kits, and hostile governments or corporations. He proposed this as a tought experiment to try to kickstart a conversation that could eventually lead to solutions implementing his ideas or, he happily admitted, those of smarter and more skilled people in this space. He spoke about the weird union of Hayek and Marx in trusting the knowledge and ability of those at the edge, with boots on the ground, over centralized authority and control. I expect a transcript and video of this talk will be up soon but he's also doing a version of it at the Long Now Foundation in San Francisco this coming Tuesday.

With both Chris and Cory, I got a chance to go to the lengthy Q&A session after their talks and to talk to them a bit about their thinking, as well as hearing them discuss the kind of stuff that doesn't really make it into a public presentation in front of a thousand or more people. In both instances, I'm reminded of what genuinely approachable and nice guys both of them are. I've interacted with both of them on and off before, mostly because of my involvement in my hackerspace or (in Cory's case) because I'm a science fiction geek. I've always found both of them to be the kind of guys that you could ask honest and real questions of and get genuinely thoughtful responses. That's always cool to find.

I also managed to give Cory and Ace Monster Toys sticker and snap a shot of him with it as well, which makes the hacker fanboy in me happy.

Update from August 5: Cory has done his talk a second time, at the Long Now Foundation in San Francisco. He has put his slides up and you can listen to the audio of the Long Now version of his talk.

Learning to Play Go Online

2012-07-14T12:51:00-07:00

I've been learning to play Go lately. Like many people, I learned the rules when I was much younger and then forgot them due to lack of play. Growing up in America, Chess has always been the traditional game of strategy played by geeky sorts. (I'll admit to being in chess club in junior high school.) Go has fulfilled that same role in much of Asia but has never taken off as much in America (or, I suspect, Europe). That said. it has a strong following these days and has been gaining popularity over the last few decades. After my childhood Chess days, I went on to role-playing games (Dungeons and Dragons) and various popular board and card games, continuing these, off and on, to this day. I haven't played Chess much in many years though I do have a set or three around the house. I never really got into Go though.

Now, here I am decades later and a bit of a fan of many things Asian, though I already was as a child to some degree thanks to my mother's love of Japanese culture, especially Toshiro Mifune. I'm a Zen Buddhist and, strangely, it is hard to be a Zen practitioner and not run into Go on a regular basis. In addition, there is a strange intersection of people who play Go and computer geeks. At least a few acquaintances of mine originally met playing Go. Because of this irregular exposure and a lifelong love of games, I've been wanting to really learn how to play.

The problem with learning Go, especially by yourself, is that the rules are deceptively simple and the game is amazingly subtle. It is easy to play bad chess and to start to get a feel for the different pieces and how they interact. You have to learn to see patterns in play by playing against a real person and getting a sense of what works and what doesn't. It is additionally difficult to play go against a computer as there aren't really good programs for it. Unlike chess, which various specialized supercomputers can play admirably well, there has not been a mastery of go exhibited by a computer program at this point. I gather that the game is a complex enough problem that computers are still relatively incapable of winning against a good go player (which I am not).

The end result is that you really need to play against people, as well as study existing games or perhaps read a few books on the matter. Since this is not an uncommon problem and people love to play, there are a variety of solutions.

You can join one of the many Go clubs, if you happen to have one in your area. Near where I live, the Berkeley Go Club has been going for a many decades. There is a regular meetup group for people who play in Oakland, as well. This provides for the ideal, which is in person play with people. (A longterm leader of the Berkeley Go Club was the original owner of the goban at the top of this post, which now resides at Ace Monster Toys.)

As a geek, though, I've found that I like online play. With some resources, you can play asynchronously, allowing each player to play a turn when ready without the other having to watch, or you can play simultaneously in a live match.

The dead simplest way to play online that I've found is Dave Peck's Go, a simple webserver based system that allows you to play with single opponent and which will email each player when it is their turn. You can specify board size, handicap, etc., which is rather convenient. Like a real mensch, Dave even has his code for this up on GitHub so you can even run your own server if you want. This just requires a web browser and email, which makes it pretty easy, and allows for people to send moves and move on to other things.

The downside of the Dave Peck's server is that it is a little clunky and only really works if you already have a designated player available (such as I did above). If you need to find an opponent because you don't have a handy Go partner (or it is frustrating to play with only that person), there are places to go to find many other players.

For the real Go action, the place to…go…seems to be Pandanet, which runs the big "Pandanet Go Server." There is a KGS Go Server as well but it requires you to run a Java client and friends don't let friends run Java in their browsers (it is probably the #2 vector for attacks on your computer and should be disabled), leaving it out for me.

The Pandanet Go Server is an IRC-like application (you can watch text stream by in the console) that mediates Go games. This server has a number of very cool things going for it. Once you make an account, you can provisionally self-assign yourself a rank. (Did I mention that Go players are ranked by skill?) Once you play a 20 games against another ranked opponent in a conforming manner, your rank is set in the system. Pandanet's server allows you to see a list of players who are connected to it, their official rank, and whether they want go play a game. This means that you can select an opponent of a suitable skill level to work on your play (or even a much more skilled player and play with a handicap) and actually advance your rank over time. You can record your games for later review and even send them to others. Additionally, you and others can observe the play of any games on the server, allowing you to watch (and review) the games of much more skilled players. All in all, this makes it an ideal environment in which to play Go if you aren't going to do it in person.

The really cool thing about Pandanet's server is that it is so ubiquitous that many people have written clients that can talk to the server. For OS X, which I normally run, there is Sente Goban, which is very nice even though not entirely open source. It is based on the Free Software Foundation's GNU Go application, an open source (cough free software) project going back over 20 years (which hasn't released a new version in three years now). Sente Goban is able to connect to Pandanet's server, allowing me to log in and play games from my laptop. Pandanet offers their own older client, glGo (or here) but the OS X version doesn't work anymore and I haven't tried the Linux and Windows versions. Their officially listed client, GoPanda is another Java app that you shouldn't run. They also offer the Panda Egg Go client for Windows.

The best way I've found to play on Pandanet is on a portable device. They have wonderful iOS and Android clients called "Panda-Tetsuki." The clients have a lovely page with links to the various versions and they are available as free applications. I've embedded their video for the iPad client below and it is a great app.

As far as basic resources for learning to play Go, there are a number of decent ones online. You can download a free copy of the introductory book, The Way of Go by Karl Baker to get you started on learning the rules and play. There is also the somewhat clunky The Interactive Way of Go. Sensei's Library also has a set of dedicated beginner's pages. "ChiyoDad" wrote a delightful blog for a few years, ChiyoDad Learns Go about his adventures learning Go as well.