Screaming Penguin

TotSP Dormancy

charlie.collins — Fri, 13 Jan 2012 01:35:01 +0000

We haven't posted anything new here in a while, for several reasons.

First we've had some server issues that have caused difficulty (like loss of article formatting) and we've decided to look for/write a new platform for blogging. We tried a few of the common ones, but they don't seem to take to kindly to importing 5K posts. So, it's a little complicated.

Second, we're using Google+ and Twitter a lot more, and that often scratches the itch. Sure we should take more time and write a more thoughtful full length article on our blogs, but, we don't get around to that too often (and TotSP used to be where everything was posted, insignificant or not).

Anyway, until we get around to really porting TotSP somewhere, look for your ususal Collins and Kebernet noise tangents here:

Charlie on G+

Robert Cooper on G+

And . . . . against all odds, we're back!

charlie.collins — Sat, 13 Aug 2011 20:21:20 +0000

A RAID mirror blew up at TotSP headquarters last week. This brought the penguin down. I had backups, and extra hard drives, so I spent a few hours on it last week. After looking at Drupal code and trying to makes heads or tails of anything logical, I ended up just kicking a big dent in the cabinet and walking away. Well, today, against my better judgement, I spent another 4 hours getting it sort of ironed out. Things are working again, but I can't upgrade the Drupal installation (that blows up). I really hate Drupal, did I mention that? I've been saying I just need to write my own blog engine on App Engine for years now, and again I'll say it, it's past time. I have all this historical content that I think SHOULD be online, but I don't have the time or will anymore to maintain a server in the basement and pay for the power and so on. TotSP is the last thing I don't have hosted in the cloud (all my other personal stuff is). A few things might not work at this point after the restore (like quotes, those came from the OLD penguin database BEFORE drupal, which I haven't been backing up since 2004 ;)), but it's mostly BAAAAAAACK! ... for now.

Git remote branch workflow

charlie.collins — Wed, 27 Jul 2011 03:14:47 +0000

I'm a relative newb to Git. I like it, but I'm sure I'm still missing a lot. Anyway, I recently had to work out a way to create remote tracking branches, because I like branches, and I want them backed up and shareable. Here's how I did it, in case anyone else might find it useful (or in case anyone can let me know what I'm doing wrong and improve my workflow). (Taken from Google Doc notes I made whilst working this out.)

Git is distributed, so there is no real “true” source. All repos are equal. A server is often used as the main/master/HEAD source of code (for instance, GitHub), this server can host repositories that are “cloned” locally. When you “clone” a remote location’s repo, using a URL, this does several important things:

Creates an alias for the “remote” URL using the default name “origin”

Creates a local *tracking* branch for each branch in the cloned repo

The default “master” branch is therefore tracked as "origin/master." Once you have a local “tracking branch” for origin/master, you may want to use branches so that you can work on new features in your code and not muck up any existing code. You may also want to create REMOTE branches, that are tracked locally, so you can also push your branch changes to the remote server (so they are backed up, and so that they can be shared with others). To create a NEW REMOTE TRACKING BRANCH: Create a remote branch Create a local branch that tracks it Make changes and commit Push to remote See what “refs” or references you have (which may help with the next command)

git show-ref

Create remote branch (assuming you have “origin” remote and “master” branch to begin with)

git push origin master:refs/heads/new_branch_name

List branches, including remote

git branch -r

Checkout remote branch locally, and track

git checkout --track -b new_branch_name_local origin/new_branch_name

Make sure you’re now on the remotely tracked branch

git branch

Do some stuff and commit/push while in the branch!

git commit
git push

When ready, switch back to master branch, and MERGE in feature branch

git checkout master
git merge new_branch_name

PROFIT!

Open Letter to Sens. Isakson, Chambliss and Rep Lewis

kebernet — Tue, 12 Jul 2011 00:34:54 +0000

I am writing to you with a sense of dismay right now. We certainly all understand the fiscal problems that face our nation at this point in time, but recent actions by the House Appropriations Committee are deeply troubling. Specifically, zeroing out the budget for the James Webb Space Telescope. I watched the end of manned space flight in the US last week with some ennui. While the Space Shuttle program has basically been a boondoggle for the entire history of the program, it is sad to see the United States back away from one area where we have been leaders my entire lifetime. Largely, though, I accept that the STS program was a waste of money. Cutting NASA's science budget, and in particularly the space telescope program is, frankly, eating our nation's seed corn. While pure science research has never been predictable in its results -- exploring the unknown never is -- it has proven itself to be the soundest investment we as a people can make, paying back dividends many times over. The Hubble Space Telescope has been a remarkable tool for pushing back the boundaries of human knowledge, and to not follow on that success with the leap that the JWST represents would be a mistake. Perhaps as importantly, the Space Telescope program represents American excellence. With the Clinton administration's abandonment of the Superconducting Super-Collider, and the imminent closing of the venerable Fermi Nation Laboratory, the US has ceded leadership in high energy physics to Europe and Japan for at least a generation. The Webb observatory represents an important foot in the door for the United States to retain some kind of leadership position in the quest to advance human understanding, lest we be shut out to also-ran status. The regular discoveries made by NASA's pure-science programs serve one other, I believe, critical purpose: they inspire our young. Astronomy is a gateway drug for learning in a way few other sciences are. Young children, as I did, can easily grasp the awe of the universe, and begin learning applied trigonometry before they even have a solid grounding in their multiplication tables. The study of astronomy then introduces them to engineering, physics, geology, advanced mathematics, and shapes them whether they become astronomers or not. I am a software engineer making telemedical systems that provide first-rate neurological care to the residents of rural Georgia, and the nation. However, without the inspiration that reading free publications from NASA as a child, learning to use my telescope, and learning about the Universe in the large gave me, I likely wouldn't be in this field I love. Most of the scientists, engineers and mathematicians you have met wanted to be an astronomer or an astronaut in their childhood. Simply ending American leadership in this field will affect us not just now, but will ripple into further erosion of STEM studies in the United State for at least a generation. While this letter might be premature in the legislative process, I urge you to support funding the Webb observatory, and NASA's science programs. As was recently noted in the paper, we spend more money air-conditioning tents for our forward-deployed troops right now than all of NASA's budget. Cutting important science programs that represent a triffling portion of the Federal budget is a short-sighted mistake of the first order. Thank you.

Don't Do It HP!

kebernet — Thu, 30 Jun 2011 00:49:21 +0000

So a number of places are reporting that HP is looking to license WebOS. Don't do it! Seriously, I think HP is way underestimated right now and their PC dominance could be a thing, but they don't need to get into licensing their OS. What they need to be doing is starting to drive specs. Get HP Touchstone on iOS and Android. Negotiate the come to jesus meeting on installable HTML5 apps with Mozilla, Google, RIM and Microsoft. Let's be serious for a moment. The mobile world being as it is right now, the indemnification of software licensing alone would eat up any profit you might have, and it isn't like HTC, Samsung or Moto are shipping profoundly better hardware than HP can on its own. The big thing is, HP needs to (a) get a tablet offering out soon and (b) beat Microsoft to the punch with touch-based HTML5 apps for PCs. THEN start licensing WebOS for Windows. Seriously, HP has a window here to take their colllective balls back from Balmer. That Windows 8 demo was interesting, but with the Palm buy, HP has a big head start. Webkit and V8 are in good shape, while IE10 is a half-assed tech preview. WebOS has package management, app store, etc. Win8 is a Powerpoint stack. But most importantly, WebOS has potential allies in Mozilla and Chrome for making installable HTML5 apps a thing right now. And frankly, Mozilla and Google could win the geeks while HP wins buisiness and "your proverbial mom" and you could all do it while Microsoft is still putting on their socks. The Touchstone stuff isn't an "end of line" tech, but it is a good start for where we need to be, and others would love to play with you. Hell, you might even get Apple on board for that and for the first time in two decades be able to dictate driver APIs outside of Redmond. Don't let the lure of quick revenue distract you from the immense soft-power potential.

Google+ -- The view from the stands.

kebernet — Wed, 29 Jun 2011 15:11:12 +0000

So I guess the buzz, pun intended, is about Google+. GOOG's boil the ocean, pun intended, social product code named Emerald Sea. Steven Levy has a huge write up at Wired about the product. He is a good story teller, but I think he misses the point. He misses a lot of stuff. He talks about the failure of Wave and Buzz to set the world on fire, but he misses WHY exactly. On Wave he says:

a thrilling demo at Google’s I/O conference introduced the social-based communications system Wave with a bang — but confusion about how to use the product dissipated the enthusiasm

And on Buzz:

itegrating some aspects of Facebook and Twitter into Gmail. Buzz’s innovations never had a chance to win an audience, as privacy flaws in the product’s initial design generated an internet firestorm

He has the kernel of it in each of these, but seems to miss the point. Wave *should* have been integrated with Gmail, and Buzz a standalone site. Why? Wave was Yet-Another-Inbox I was supposed to check. That is crap. I already have an inbox. Getting change notifications is not the same. Wave should have become the killer GMail feature. Moreover, it needed better integration with Sites, etc, so content developed in Wave was portable. Moreover, Wave for Domains needed to drop immediately. Buzz failed because it added crap I don't want in my inbox TO my inbox. Facebook and Twitter aren't "inboxes". They are streams of stuff that you can pay attention to for a little while and then ignore. Having to go in and "mark read" crap in buzz was idiotic. Buzz was stream of conversation at a cocktail party. You don't want to have to check off that you tracked every thread of conversation around the bar before you leave. Google had good products in each of these, but a fundamental misunderstanding of how people wanted to interact with the data. It looks to me like they have done it again with Google+. Google plus really seems like a hodge-podge of features rolled into a single UI. That isn't bad. I have yet to see exactly how it relates to all the other Goolge properties (Talk, Picassa, Reader, Contacts, +1, Buzz) that it seems to encompass. My fear is it is the beginning of Google "Doing the Yahoo!" -- where they have five different properties that do about the same thing with slightly different branding and no interoperability. The TechCrunch coverage definitely has that smell. This, however, is not why I am initially negative on G+ though. I am down on the service because it provides a collection of things people don't want. Circles? Are you fucking kidding me? You know how many people use Groups on Facebook? Like 3, and there were likely 6 people on the dev team for groups. Lists on Twitter? More than three, but few. Groups in Google contacts? I am going to go with nearly none. The fact of the matter is people don't WANT to organize yet another bloody buddy list. They don't even organize the ones they have. But the thing that kills me about Circles is how anti-Google it is. What was the big innovation of Google? Making search so useful that directories (like Yahoo!) became passe. What was the big innovation of Gmail? Adding great search to mail so I don't have to sort crap out into folders like we all did before hand. Oh wait, then they added Priority Inbox, which might be the single greatest innovation in communication since the pager. These are all great. You know why? Because they fit with Google's mission statement: "Google’s mission is to organize the world‘s information and make it universally accessible and useful." Circles has changed this to "To get our users to organize their own shit." Look, organizing contacts would be nice, but I don't want to do it. Those of us who live online are usually dealing with WAY PAST the Dunbar number of people across Gmail/Talk, Facebook, Twitter, Last.fm, Work, etc. It would have been much more interesting if Google could just start watching my interactions with my contacts over these various properties and start autogrouping and subgrouping them on its own. Congrats. You hired one of the most famous UI people in the world and got him to build a beautiful interface I have absolutely no desire to use. Huddle and Hangouts? OK. Whatever. First, why were these not just rolled out as Google Talk features? Huddle is certainly useful, and seemingly timely since Apple has their BBM clone coming out. But it is still just a chat room. Those have been around forever. Hangouts? Look, multi-party video is, again, a good thing to have, but "Hangout?" Are you expecting me to, what, turn this on while I am at my desk so people can watch me go about my business? Huddle -- the *CHAT* product is the "Hangout" -- where you maintain presence all through the day. The *VIDEO* product is the "Huddle" where we are all heads in at the same time. Seriously fucking epic name fail. What else is here? Well, Yet-Another-Photo-Thing. "Sparks" or "Interests". I dunno exactly what that means, aside from "Tell us what ads to sell you." That might have been more interesting as a merger of Facebook "Fan Pages" with the functionality of Google Groups. All in all, I liked Wave a lot, and I still use Buzz, but I don't use it in Gmail, I use it via Reader which is more in line with the ephemeral model it should have been. I honestly can't even seem myself WANTING to use G+ -- and I am a fanboi.

Brilliant.

kebernet — Tue, 28 Jun 2011 15:41:07 +0000

The Daily Show With Jon Stewart

Mon - Thurs 11p / 10c

Oh, for Fox Sake

www.thedailyshow.com

Daily Show Full Episodes

Political Humor & Satire Blog

The Daily Show on Facebook

PokerFun.tv on Engadget

kebernet — Sun, 26 Jun 2011 22:57:51 +0000

Just because Chuck hasn't posted it yet. WTG MOVL.

Lists are an incredibly lazy way to fill blog content...

kebernet — Wed, 22 Jun 2011 01:24:34 +0000

... but NPR wants a list of 100 greatest Sci-Fi/Fantasy books. Fantasy sucks. In general decending order: 0. 1984 - Orwell (Really, this should be before #1) 1. Brave New World - Huxley 2. The Time Machine - Wells 3. 20,000 Leagues Under the Sea - Verne 4. Frankenstein - Shelly 5. Dune - Herbert 6. Neuormancer - Gibson 7. The Handmaid's Tale - Atwood 8. Foundation and Empire - Asimov 9. Rendezvous with Rama - Clarke 10. The Martian Chronicles - Bradbury 11. The Difference Engine - Gibson and Sterling 12. The Hitchhikers Guide to the Galaxy - Adams 13. Flow My Tears, The Policeman Said - Dick 14. The Watchmen - Moore 15. Stranger in a Strange Land - Heinlein 16. Ringworld - Niven 17. A Clockwork Orange - Burgess 18. I, Robot - Asimov 19. Contact - Sagan 20. Breakfast of Champions - Vonnegut 21. Ender's Game - Card 22. The Space Merchants - Kornbluth 22. Fahrenheit 451 - Bradbury 23. Pattern Recognition - Gibson 24. A Boy and His Dog - Ellison 25. The Andromeda Strain - Crichton 26. Akira - Otomo 27. A Journey to the Center of the Earth - Verne 28. The Cryptonomicon - Stephenson 29. Oryx and Crake - Atwood 30. Flowers for Algernon - Keyes 31. Slaughterhouse Five - Vonnegut 32. God Emperor of Dune - Herbert 33. At the Mountains of Madness - Lovecraft 34. The Gods Themselves - Asimov 35. Islands in the Net - Sterling 36. American Gods - Gaiman 37. Rainbow's End - Vinge [spell corrected] 38. Heir to the Empire - Zahn 39. The Gunslinger - King 40. Childhood's End - Clarke 41. Life, the Universe and Everything - Adams 42. The Sirens of Titan - Vonnegut 43. The Road - McCarthy 44. Solaris - Lem 45. Red Mars - Robinson 46. Flatland - Abbot 47. Old Man's War - Scalzi 48. Caves of Steel - Asimov 49. The Postman - Brin 50. Starship Troopers - Heinlein You can fill in 50 crap fantasy books from here. While Kid-Lit is out of bounds for the NPR poll, I give a special #20 Award to: 20. A Swiftly Tilting Planet - l'Engle Thanks for playing!

Military-Industrial-Congressional Complex

kebernet — Mon, 20 Jun 2011 03:55:11 +0000

This via mesh:

It is not democracy that keeps western nations at war, but armies and the interests now massed behind them. The greatest speech about modern defence was made in 1961 by the US president Eisenhower. He was no leftwinger, but a former general and conservative Republican. Looking back over his time in office, his farewell message to America was a simple warning against the "disastrous rise of misplaced power" of a military-industrial complex with "unwarranted influence on government". A burgeoning defence establishment, backed by large corporate interests, would one day employ so many people as to corrupt the political system. (His original draft even referred to a "military-industrial-congressional complex".) This lobby, said Eisenhower, could become so huge as to "endanger our liberties and democratic processes". I wonder what Eisenhower would make of today's US, with a military grown from 3.5 million people to 5 million. The western nations face less of a threat to their integrity and security than ever in history, yet their defence industries cry for ever more money and ever more things to do. The cold war strategist, George Kennan, wrote prophetically: "Were the Soviet Union to sink tomorrow under the waters of the ocean, the American military-industrial complex would have to remain, substantially unchanged, until some other adversary could be invented." The devil makes work for idle hands, especially if they are well financed. Britain's former special envoy to Kabul, Sherard Cowper-Coles, echoed Kennan last week in claiming that the army's keenness to fight in Helmand was self-interested. "It's use them or lose them, Sherard," he was told by the then chief of the general staff, Sir Richard Dannatt. Cowper-Coles has now gone off to work for an arms manufacturer. There is no strategic defence justification for the US spending 5.5% of its gross domestic product on defence or Britain 2.5%, or for the Nato "target" of 2%.

Reminds me of George Washington:

Altho' a large standing Army in time of Peace hath ever been considered dangerous to the liberties of a Country, yet a few Troops, under certain circumstances, are not only safe, but indispensably necessary. Fortunately for us our relative situation requires but few. The same circumstances which so effectually retarded, and in the end conspired to defeat the attempts of Britain to subdue us, will now powerfully tend to render us secure. Our distance from the European States in a great degree frees us of apprehension, from their numerous regular forces and the Insults and dangers which are to be dreaded from their Ambition. But, if our danger from those powers was more imminent, yet we are too poor to maintain a standing Army adequate to our defence, and was our Country more populous and rich, still it could not be done without great oppression of the people. ... Were it not totally unnecessary and superfluous to adduce arguments to prove what is conceded on all hands the Policy and expediency of resting the protection of the Country on a respectable and well established Militia, we might not only shew the propriety of the measure from our peculiar local situation, but we might have recourse to the Histories of Greece and Rome in their most virtuous and Patriotic ages to demonstrate the Utility of such Establishments. Then passing by the Mercinary Armies, which have at one time or another subverted the liberties of all-most all the Countries they have been raised to defend, we might see, with admiration, the Freedom and Independence of Switzerland supported for Centuries, in the midst of powerful and jealous neighbours, by means of a hardy and well organized Militia. We might also derive useful lessons of a similar kind from other Nations of Europe, but I believe it will be found, the People of this Continent are too well acquainted with the Merits of the subject to require information or example. I shall therefore proceed to point out some general outlines of their duty, and conclude this head with a few particular observations on the regulations which I conceive ought to be immediately adopted by the States at the instance and recommendation of Congress.

The Game Loop (Or: Why you fucking video game people can't deal with multi-core)

kebernet — Sun, 19 Jun 2011 04:00:51 +0000

Recently saw this come up on Reddit. Dear Game People: STOP FUCKING THINKING LIKE IT IS 1987. While the rest of the world is finally moving to Scene Graphs for plain-old-apps, you people are writing code like it is 1987. I understand it is easy to think this way, but it is time to face up to the new world of multi-processor CPUs and dealing with concurrent state. You need to be handling input on an event-based system, manipulating your data model, then in another thread translating that manipulation to a new scene graph and publishing it. Your "network" code then has the same fucking interface to the game data that your UI code does, and -- wait for it -- it is just another client. Moreover, all of this is non-blocking and FPS issue fix their fucking selves. Look, I get most of you aren't CS majors, and you have been doing this crap the same way since the 70's, but your games are starting to suffer. Getting your network code nailed to the same CPU as user input is lame. Even lamer: I get that doing your unified state machine is easy, but seriously, start doing distributed resolution and sync. Capture event times in your model and resolve conflicts based on a fair time-sync system. Computers have changed. It is time to put on the long pants.

"Software is an entropic system whose arrow of time flows in the direction of failure, aided and abetted by human bullshit."

kebernet — Sun, 19 Jun 2011 01:51:08 +0000

Love "A Platitude on Software Failure".

But don't kid yourself that the high-hanging fruits of the technology tree—strong AI, for example—will be made manifest by a society more excited by the Facebook social graph than by feeding its children.

Web vs Native: Where we fell

kebernet — Sun, 19 Jun 2011 00:42:48 +0000

The web vs native app thing is coming around again. Tim Bray has a great write up and object lesson in response to the buzz from TechCrunch and others. I thought I would sound off on this. I am of a few minds on this. My current employer, I would say we are SQUA in the "Web" side of this. We have done a lot of work to provide a quality web app for a number of reasons: first there is zero install for a lot of use cases. While most of our work is done at desktop computers, having a system that is accessible with no additional software is important. Indeed, one of the "tales" of our company (telemedicine) is of a doctor performing a consult from the manager's office at a Best Buy while on vacation. Sure we aren't talking about mobile, but I don't see that as an important distinction right now. Our app is decidedly unsuited to running on a phone, but tablets are clearly in our sights. In spite of this, however, we are building a native desktop app. This desktop app mostly just runs Chromium Embedded and provides hooks to control video conferencing controls to the native APIs. There are lots of things involved here, but the short version is, we simply COULDN'T get the support we needed through a stock web app. While our "native" desktop app is barely more than a web browser++, controlling Pan-Tilt-Zoom cameras, installing a virtual printer driver, etc, were things we couldn't make happen from a basic, or even an advanced web app. So basically, we have a "PhoneGap" app for the desktop. Our app, however, inherently needs connectivity. Yes, HTML5 allow for better off-line support, but honestly the packaging matters, both for our desktop app and mobile apps. Tim is certainly right to point out the TripIt app, and the lack of a need for a mobile app -- with one qualification. TripIt could build a perfectly good PhoneGap version of that app, but running it out of a browser would blow. Why? The browser's chrome and UX really kind of suck. More importantly, for developers, it encourages you to fall back to "traditional" web development models, rather than build a "Web Client" app. Secondly, off-line access for that kind of app matters A LOT. You don't want crappy network coverage in O'Hare to prevent you from seeing your flight information. Again, you "can" do that with HTML5 apps, but even building the "web" version makes developers lazy about thinking that way. Building a PhoneGap app doesn't. As much as I love PhoneGap and hate AppCelerator, they are really important, like our client is, for making sure you can do those things you need to do on a device. Access to local data is a big one. Not HTML5 storage, but Contacts databases, non-interactive mode filesystem ops, etc; these matter as much if not more than simple "Off Line Data Storage." Now, while I think native-wrapped web is a win, there are certainly other areas where there is a potential for "win." For basic apps, TripIt is a good example, you will write a LOT less code. Certainly for the "Enterprise" market, this is critical. Using things like GWT on the client or Node.js on the server means you can get a whole lot more code re-use through all layers of your application. Even for more "product" type companies, having unified code bases means that there is less bifurcation of features and options between different platform clients. The huge variation in quality and feature support of the many, many Facebook applications speaks directly to this. As an Android "guy," it is frustrating to flip back and forth between the mobile web app and the native app depending on which is "ahead" at any point in time. Right now the mobile app is light years head of the Android app, but a well build native wrapper could provide all the Android-ish hooks to Contacts, Calendaring, notification, etc that the native app does, while still providing the improved interactive UI that the mobile app currently does. I won't speak to games. Charlie could address that better than I could. But "web only" right now fails in a lot of ways. Even in cases like Google Reader or Facebook where the web app is better than the native for most uses, the native version provide important features you just can't get from the web.

Microformats Considered Inferior

kebernet — Sun, 05 Jun 2011 04:23:14 +0000

Now, I have written on this topic before, some five years ago. However, I was having a discussion with Kevin Marks on Twitter early this morning and I thought the topic bore revisiting. While those of us up early or late in the US were enjoying live discussion of the Federated Social Web conference the topic of Microformats came up. Kevin specifically pointed me to "Namespaces Considered Harmful" on the Microformats.org site. It is actually a good read, but it has a number of flaws. Outlining the arguments, lets look at the major topic headings: "Namespaced Content has Failed". OK. I'll bite on this. It isn't widely supported. Or is it? XHTML certainly never set the world on fire, so Namespaced content on the web never got that underlying support. However, the more important reason, as I noted half a decade ago, that Namespaces on the web never took off was less about XHTML, and more about the lack of CSS with Namespace support in MSIE. The decade long nap Microsoft took in the development of their browser made wide scale adoption practically impossible. Let us look at the subheadings under "failed" though: "namespaced content is not well supported". Excluding MSIE I would say it is very well supported, and MSIE is finally showing signs of life again. But the whole point of Namespaced content wasn't about the "browser". It was about everything else around the browser. Certainly you can expect most modern music applications to support RSS or Atom with the Yahoo media extension. RSS/Atom feeds with contacts, calendaring, photographs, etc are widely supported by all sorts of applications in their relevant spaces. PubSubHubub, while not huge, is showing all kinds of life, especially around things like Salmon (comments) and ActivityFeeds, and finally as a good event-driven API for webbish applications. To say namespaced content is not well supported is to limit yourself to only to web browsers. The problem is, the world of Microformats is confined exclusively to web browser. Sure you can get plugins for your browser to extract a vCard, but can you point your contact manager at the web page? Wouldn't it be nice if the homepage for Wait Wait Don't Tell Me *was* the podcast? If they could embed MediaRSS in it and everything would just continue to work? Let's move on to the next major topic, "namespaces for content are a negative". Why? "namespaced content discourages interoperability of data". Well OK. And how does a Microformat address this? hAtom, held up as an example, expects you to use "entry-title". What about "title", "subject" "topic"? Yes you can overload a class list, just as you can nest multiple namespaced content. However, GOOD interoperability requires standardization for common things. Again, MediaRSS, Slash, and Dublin Core have ruled the roost here, but the interoperability they have given us isn't about Namespaces vs Microformats, it is about widespread adoption. Basically, this is a moot point. Microformats are no better or worse for this. "using namespaces cost a lot of time". This argument is made with a one off snarkly line, "I wonder how many hours in my life has been wasted looking up namespace URIs for copying and pasting". Well, shit. How many hours have I spent remembering the differences between CSS implementations and, more recently, slight differences in -webkit-* and -mozilla-* names? Is CSS considered harmful? Why not ask why there isn't better tooling around this? Heck ROME to which I have devoted a lot of time, takes care of tons of that, plus converting between competing specifications for licenses, embedded content, event entire language specs. Write you app to one spec, and support everything else too. Programming is hard. Let's go shopping. Next, "bound prefixes are an anti-pattern". This section goes on with a bunch of babbling arguments about how bound prefixes require you to do stuff the way the prefix outlines. Sure, I am sorry that you have to live with the requirements of the spec to get the benefits of using the spec in your app. Until we all have semantic natural language processors as part of our standard toolkit, though, live with it. Honestly, all of these arguments remind me of the static vs dynamic types in languages. There is no need to fight this particular Holy War. This is my favorite, "namespaces cause dogmatic noise". Sweet Zombie Jesus. Namespaces cause noise, and might get multiple definitions in a document. Woo hoo. The answer? Make sure that, in addition to the data format you might want to use, your application can handle all the noise that comes along with a modern HTML page -- a mish-mosh of noise around the data to say the least. Let us not kid ourselves, unless you are looking at a demo from CSS Zen, ALL modern web pages are a noisy mess already. Even those CSS Zen people could style a Atom+Namespaces page to look as good as the HTML if they simply had the client-side rendering capability. At least with Namespaced content you can easily collapse text inside the area of interest to extract or generate data without having to have the fully syntactic support of a web browser available to you. But even looking at the new Schemas.org site, (Microdata, not Microformats) talk to me about "dogmatic noise" in " &ltdiv itemprop="director" itemscope itemtype="http://schema.org/Person"> Yeah. That is a lot better than < m:director><p:name >. And a lot clearer, because, you know, there won't be anything else you might want there or extraneous data. Finally, "non-namespaced techniques have been succeeding". To circle back around, this is only true if you count stuff that happens inside the web browser. Sure, if I want to pull in a bit of JavaScript that does syntax highlighting, I can do <code class="php"> and it can deal with it. Show me something that isn't a browser plugin that can deal with Microformatted HTML. Outlook? iPhoto? Windows Media Player? Seriously. One example. Moreover, show me a web application in the wild that can use a foreign HTML page as a data source based on Microformats. Indeed, the world of the web has now given Microformats a complete side step and gone directly to standardized JSON and /json+* media types. In conclusion, Microformats are a toy for a certain group of webbish Geek. To say they have seen any success outside this user base is a lie, while namespaced content is now widely supported, and used, by tons of applications. Admittedly, namespaced content in HTML has been a failure, but that is almost solely at the hands of Microsoft.

Review: Ona Union Street bag (onabags.com)

kebernet — Sat, 04 Jun 2011 06:06:08 +0000

As some people may or may not know, I have been on a mission from god for a new bag for a number of years now. I had a bag I had carried for decades that was stolen a number of years ago. Since then, I have carried freebies and defaulted back to a (surprisingly durable) messenger I got from The Gap, of all places, when the freebies broke. I finally found a new bag, and it is just shy of orgasmically good. The Union Street from Ona Bags is almost perfect. While it is sold as a camera bag, it meets nearly all of my requirements: heavy gauge, waxed cotton canvas. It is structurally sound. Proper leather accents and antiqued brass fittings. As for quality of construction, feel, and features (including two mobile device pockets, and four memory card slots) it is pretty damned awesome. You might think it just a little pricey, but it is worth every damned penny. The "camera bag" nature shouldn't fool you. If, like me, you need a laptop bag with organization for phone, tablet, a plethora of cables and whatnots, the movable barriers in the front pockets are perfect. The problem with too many laptop bags, IMHO, is they are centered around being brief cases with a laptop slot. Screw that. I need a slot for my laptop brick, a slot of misc cables (RJ45, HDMI, etc) and pockets for fobs. There are flaws. Honestly I wish the slots geared for SDCards were larger. Large enough to hold oversized USB fobs (Flash, WiMAX, etc). It has pen pockets, but they are too small to fit a Sharpie in. The magnetically closing back-side flap, though, is perfect for you tablet, and the edge pockets will take a mini umbrella perfectly, depending on your handedness. The one thing I would call a critical flaw is the size. For a 13" MacBook, it is ideal. They sell it as good for 15" laptops, but my Dell 15" work machine seems to almost over-extend it. There is no chance you are going to get a 17 or two 15s in this puppy and still close it. Still, I have to say, I feel like I have a nice bag for the first time in years. Rating: 5.1 penguin heads on our scale of 5.2.