News You Can Bruise

We've just secured a multi-year grant to expand the project, and we are hiring up from two developers to eight. We are quadrupling the size of our development team.

This is a really satisfying job for me because I'm making life substantially better for people who aren't already well off. If you like that prospect, if you like what I say in the "Enterprise Media Distribution" talk, and you want to work on this project, you should apply for one of these position by sending your resume to info@librarysimplified.com.

I'm going to link to the job listings in a minute, but first I want to make it real clear that we put up these listings largely to have entry points into the HR system. As the team lead I'm not concerned with counting how many terms on your resume match terms used in the job listing. We need two Android developers and four people to write server-side code and HTML and Javascript. I don't think we need a team made up entirely of Senior Developers. Other skills might be more important.

For instance, we need someone with devops experience. We'll be dealing with e-commerce, cryptography, and machine learning—all things I know little about. We don't care if you have a CS degree, but if you have a Library Science degree or have worked in the publishing industry, that would be useful. We have big collections in Spanish, Chinese and Russian, but nobody on our team reads those languages. Stuff like that.

With that in mind, here are the job listings:

• Senior Mobile Developer (Android)
• Senior Full Stack Applications Developer
• Senior Web Applications Developer.

As you can see if you click around, getting into the HR system to formally "apply" for these jobs requires filling out a really long form. (Update: and now these links don't even work anymore because the jobs got shifted around.) Instead of doing that, send your resume to info@librarysimplified.com and we'll only ask you to fill out the form if we want to bring you in for an interview.

All these positions are in New York City, in the big building on 42nd Street with the lions. This is a project funded by grants, and the salaries we offer are not competitive with Facebook or Goldman Sachs, but they are competitive with other nonprofits. The benefits are good. This is not a job that ruins your life. It's 35 hours a week and you get four weeks of vacation per year. I work from home about one day a week. Send me email or leave a comment if you have any questions about benefits.

• Manufacturers of polyhedral dice.
• Custom-printed toilet paper. "What do you need your tissues to be? We give you the satisfied answer now!!!"
• BOPP Tape Slitting Machine.
• Industrial food processor.
• Locate table for changing babies? You need Fully-Auto Baby Diaper machine.
• The Guangzhou Junyue Advertising Co. makes those company-name signs that get mounted on the sides of buildings.
• New York City Subway USB flash drives. (That company also has London, Beijing, Tokyo, Singapore, Taiwan, and Hong Kong versions.)

It's a little far-fetched to expect an ARPANET reference in a kids' movie released almost immediately after the ARPANET was created, but the movie does take place at a college, and I figured it was possible someone had gone to Stanford, talked to some computer people to see "what's up with computers these days", and gotten the phrase "Interface Message Processor" or something to use as a bit of set dressing.

After all, moviemakers today routinely employ science advisors to tell them that what happens in the screenplay is impossible because it violates the second law of thermodynamics; the moviemakers then decline to totally rewrite their screenplay and instead add a little scene of someone shouting "We need more thermodynamics!" I wasn't expecting accuracy, but I was hoping for a throwaway reference on a chalkboard.

But, in point of fact, Wikipedia is wrong: this movie contains absolutely no references to the ARPANET. [0] In fact, it barely contains any computer. The computer breaks about fifteen minutes in, Kurt Russell gives himself superpowers trying to fix it, and we never see it operational again. The rest of the movie focuses on Kurt Russell's new computer-like ability to scan and regurgitate a wide variety of data.

In an accidentally realistic twist, the Russell character's powers don't make him any smarter—there's no question of using him to solve the world's problems—he just knows a lot of facts. (He's also good with languages, and there's one line that claims he's solved the Chinese Room, but they don't go anywhere with this. They show him going to the UN and palling around with all the delegates, but it's not like they're asking him for advice.)

The movie also deals in an antiseptic Walt Disney way with topics like student demonstrations (misleadingly played up on the movie poster), but it all takes second place to the dean's rivalry with the dean of the nearby state college, a numbers game run by the always-enjoyable Cesar Romero, and the use of a superhuman with the powers of Wikipedia to win a trivia quiz show.

Why would someone watch this movie and think it had some reference to the ARPANET? I have no idea. I find it likely that this movie was copied into the Wikipedia entry from a more general list of old movies about computers, like this one. In the movie there's one computer component that superficially resembles an IMP, but 1) it's not an IMP, and 2) the H316 hardware used in the IMP was also used in other minicomputers, including the incorrectly-maligned Kitchen Computer, so even spotting something that looks like an IMP proves nothing.

Someone with more hardware knowledge might know what computer was used as a prop in the movie, if it's a real computer at all. Judging from glimpses of the labels on buttons, I think it's a mishmash of equipment from the early 60s that was stripped and rewired to blink randomly.

I did find one interesting bit of window-dressing: at 10:30 there's a partial shot of what looks like a cheat sheet for a computer's instruction set or operating system. It's called "POS", the instructions are divided into a "Physical Level" and a "Logical Level", and they include "EXCP", "WAIT", "OPEN FILE", "CLOSE FILE", "GET FILE", and "PUT FILE"--all of those take an argument (?) "CCB-NAME". Since the ISO networking stack has a "physical layer", it's possible someone glimpsed this card, thought "Aha! The Internet!" and rushed to Wikipedia or whatever source Wikipedia is based on. That's the best I can come up with.

If I may damn with faint praise, the first fifteen minutes of the movie are fun. The opening credits are great, with a corny sitcom-theme-style opening song (which describes a much better movie) and visuals that combine colorful 60s geometric design with computer imagery like punch card chads and reel-to-reel tape.

The first scene features a professor played by William Schallert (who'd just played the Federation bureaucrat in "The Trouble with Tribbles") trying to convince the antediluvian dean that dropping $10k on a computer (about $60k in today's money) would be a great investment for the school.

Meanwhile the loveable, nonthreatening youth of the college are eavesdropping on this conversation via transitor radio, and check this out--Kurt Russell is eighteen years old when they're filming this movie. The other kids are in their early twenties. I've been watching all these old B-movies that cast twenty-four-year-olds as high school students, and by comparison these college-age kids look too young to be in college! It's insanity.

Anyway, the kids convince local businessman Cesar Romero to donate an old computer to the college, he does so (apparently without wiping the tapes he's been using to computerize his numbers racket), there's a great scene where they set up the enormous computer, and then an interesting scene where the computer's capabilities are demonstrated (pretty realistic, by the standards of the rest of the movie). In a very strange twist, the computer breaks when the professor tries to get it to run a twenty-year-old program off a tape. (That would be a program from 1950--like a friggin' UNIVAC program or something. Supposedly this computer's twenty years old.) That's pretty much the end of the good part.

Of all the movie's inaccuracies, the worst comes at the end, when the guy who was in "The Trouble With Tribbles" reenacts the first scene with the dean, except instead of a computer, he's trying to get the dean to drop a few grand on another technological wonder, the electroheliospectrograph. Apparently computers are just a fad for this guy. Kurt Russell reprised his bland college-student character in two sequels, which employ more traditional SF gimmicks like invisibility potions and strength serums. If I'd been in charge I would have made a new stupid college computer teen comedy every few years as a kind of marker to track society's feelings towards computers through the 70s and 80s.

I watched the movie with Pat, 1) to make watching a bad old movie bearable, 2) to have another pair of eyes looking for ARPANET references. Pat's opinion: "They should do a sequel starring Kurt Russell as he is now."

The movie was remade with Kirk Cameron in 1995, and it's possible that movie is the one with the reference to the Internet, but I'm not really interested in something that late.

[0] I got a used DVD of the movie, and there were a few seconds of skipped video. It's possible the elusive ARPANET reference was in there. It's very very unlikely.

Some Moore's Law style inflation shows just how impressive: an equivalent amount of data today might be 370 petabytes. They must have had a whole tape library devoted to that weather data.

There are a few new photos at the end of the gallery, which Beril took and sent to me. But the big draw is all those awesome old computers, many of which sent sent Kevin into a fit of murderous rage. Enjoy it again--it's been a year, do you really remember that there was a computer called the Gandalf? Looking at these pictures the day after Christmas should become a damn tradition.

Today's 2009 gallery is the trip I took with Alexei to the Met earlier this month. You already saw my photos of "Insects, Birds, and Beasts", but now witness my second attempt to get decent-quality photos of all my most coveted pieces in the Met. (My first attempt will be showing up later this week as a 2008 gallery.) Highlights of the highlights: Jackson Pollock's "Autumn Rhythm" (I also took closeups), three of Florine Stettheimer's four crazy 1920s consumerism-satire/love NYC paintings[0], and a shot of an ink cake. (More shots of ink cakes coming later this week, but they're not as well-lit as this one. But this one is kind of blurry, so oh well.) And, of course, a photo of me with "Piece of Cake". If only there were a photorealistic ink drawing called "Piece of Ink Cake", it would be the ultimate Met exhibit.

[0] I can't find where the fourth one is; I'm assuming private collection.

In the tradition of mixing awesome and boring Christmas presents, today's 2008 gallery is the consumer electronics equivalent of tube socks: cassette tapes! Not the generic tapes from the 1990s, but old cassette tapes, from the 70s and early 80s, when cassette tapes had brand names like "Sears". Tapes with awesome slipcases and weird slipcase linings. How do you know it's the 70s? Two-tone cassette tape. Oh yeah.

It's not in dispute that the Honeywell H316 "Kitchen Computer" is a really stupid idea. But I do think that the pedestal version of the H316 was not designed or seriously intended to be used in a kitchen. I propose that the pedestal H316 was originally designed as a desk, something you'd sit at, using the "cutting board"/"counter" as a workspace, and that the kitchen usage was dreamed up for the Niemen Marcus catalogue as a way to sell other things--or, less likely, was a desperate gamble to sell a computer in a form factor that didn't make sense. It's possible that some pillar H316s were sold, even if no "Kitchen Computers" were.

The first microcomputer, the Altair, was programmed with switches and had no user interface other than its lights. To us it seems barely plausible that in the appliance-crazed 60s, Honeywell would stupidly try to sell that kind of interface to the home. But in a minicomputer those switches and lights are just the control panel. They're used to load in the operating system, so you can use a teletype or tape reader for your real work. A teletype next to the actual computer was a common sight in contemporaneous advertising.

The H316 worked that way. No one would do all their work through the switches, which is why building the H316 into a desk was a bad idea. But putting a teletype next to the H316 in the Nieman Marcus photo shoot--a setup that might actually have worked--would have spoiled the 2001-esque aesthetic Honeywell and Nieman Marcus were projecting. Today you could build a real computer with that form factor and no teletype, so we look at the ad and think they were trying to use the control panel as the user interface.

So: not only did the Kitchen Computer sell no units, it had no existence outside the Nieman-Marcus catalogue. The point of the catalogue entry was to make executives puff on their stereotypical pipes thoughtfully and then buy an H316 for the office. Everyone else would shudder at the cost of the computer and just buy that dorky flower apron.

Here's some questions I don't know the answer to that could confirm or disprove my hypothesis. First, I suspect that the "counter" of the pillar H316 is much lower than a kitchen counter would be; more in line with something you could scoot an office chair under. This is born out by the N-M photo, Val Henson's hilarious photos, and my subjective remembered experience, but I don't know for sure. Here's a picture of a chair beneath the computer. My point is that no matter how stupid an idea a Kitchen Computer is, if you were designing the casing for one and you decided to give it a "counter", you'd put the counter at the height of a kitchen counter, not the height of an office desk.

Second, I don't know what's in the back of the pillar H316. In the continuing saga of me not looking closely enough at things that I later realize are problematic, I didn't look at the back. Is there a place to connect a teletype and other peripherals?

Third: what did the H316 cost through regular channels? I've seen it described as "the first under-$10,000 16-bit machine". Does the $10,600 cost in the Niemen-Marcus catalogue include a cheap teletype to go with the two-week programming course?

Fourth, are there any manuals for the pillar H316? Any idea of how all those alleged recipes were stored? Paper tape? They sure wouldn't fit in the H316's core memory.

Bonus clearing-up: it has been alleged that the Kitchen Computer had a compiler for a language called BACK. There is no such programming language. Which itself is suspicious, because most short words have been used as the name of a programming language. At some point someone made a joke about FORTH (the first FORTH implementation happened on the H316) and the factoid was passed along by someone who didn't get the joke.

An interesting coda: the H316 was also used to build the Interface Message Processor, first machines on the ARPAnet. Check out the lower IMP control panel: it's got the same controls as the Kitchen Computer.

PS: I'm gonna go out on a limb and allege that the Kitchen Computer is a sham. It was never intended to sell, it was just a cross-promotional thing to get attention for Honeywell. Someone realized what a catastrophe it was to design a computer with the form factor of a desk in 1969. People in 1969 don't sit at desks and twiddle binary switches on the computer panel, they sit at teletypes and type and don't worry about where the computer is physically. So the paper-shuffling part of the desk was hastily renamed a "cutting board" and you had the Kitchen Computer.

There was probably no recipe software or digitized recipes. If they were serious about selling a Kitchen Computer they'd sell you a teletype as well. If you tried to use the "cutting board" to cut vegetables you'd scratch up an irreplacable part of the computer and probably cause a short circuit. You'd have to be stupid to design a Kitchen Computer that way. But you wouldn't have to be stupid to pass off a failed form factor as a Kitchen Computer for publicity's sake. Look at the ad: "If she can only cook as well as Honeywell can compute." Not "this computer", "Honeywell". It's a high-concept ad for Honeywell.

So there's a bunch more photos on Flickr, with more to come. New photos start here.

The big draw was the difference engine, which is unfortunate because in a few months Nathan Myhrvold will arrive Mephistophelian at the stroke of midnight and whisk the difference engine away to his living room. And then the Computer History Museum will have a rotating exhibit (currently about chess), an "Innovation in the Valley" exhibit that seems provincial despite having amazing artifacts like Engelbart's mouse and the Apple I, and the big room full of computers. You think it's ridiculous that I'm complaining about free admission to a big room full of computers, but if you've seen the outside of this building it looks like the office-park equivalent of the Guggenheim. You expect more.

Anyway, the difference engine was great, and now I understand how it works. I originally thought the point of it was to find the roots of polynomials and I couldn't imagine how it represented complex roots. But no, it just cranks out the values of polynomials for ever-increasing values of x.

More pictures from the room full of old computers coming eventually. I just wanted to put up enough that I'd get to the Chadwick Magic-Brain Calculator for Susanna and John.

Well, it didn't take me long to get up and search for those commercials online, where they now live. They're surprisingly similar to Douglas Adams's much more enjoyable Hyperland. In fact, I think Steven Levy was thinking about Hyperland when he described some aspects of Knowledge Navigator.

Today's cool thing is Ruby's midilib, the... wait for it... Ruby interface to MIDI files. I did a recipe that shows how to visualize data as a MIDI file, and how to compose algorithmic music (sample composition).

While learning about midilib I came up with an observation about MIDI, which is that it's good for accurately representing what happens when you play a real instrument. It's kind of frustrating and unnormalized if you need to programatically compose music. I ended up taking a tip from the excellent MIDI Scripter project, and showing how to write a little wrapper on top of midilib that makes composition easier.

The experience gave me a huge number of ideas for musical programming projects. With my near-total lack of theoretical training, I will be unencumbered by traditional notions of "harmony" and "taste". For one, I plan to get into fractal music, though probably not the way the author of that web page defines it. Hard to tell, with that color scheme.

Second, Josh has started a reading group for The Art of Computer Programming. His aim: use group dynamics and the fear of failure to make the members of his pact keep reading the thing and working the exercises. I always knew peer pressure was a technology capable of substantial non-infringing use.

Update: I can't believe I forgot the XML-RPC API I wrote for Downhill. It doesn't really meet my criteria either, because it's not that useful anymore. The Weblog Ecosystem data it depends on is about 2 years old and it's not being collected anymore so I can't update it.

Hmm, Waypath and Meerkat look like they might work.

It turns out that Amazon recently introduced that exact feature, so good for them. Unfortunately since it's part of their Search Inside The Dang Book initiative I doubt there will ever be API support for it. "You let hackers Search Inside My Dang Book automatically?"

One of the reasons we did the Metaweb as a separate site, and not part of the Wikipedia, is that we want users to be able to express their personal point of view, if they wish, and also to be able to create authored content. Neither of these is possible on the Wikipedia.

Most likely I misleadingly took the "replace Wikipedia" thing out of context from someone's list of goals for Metaweb. Unlike the mainstream media, I am willing to take responsibility for my attempt to take things out of context and sensationalize. Unfortunately, I have yet to progress beyond couching my taking of responsibility in self-aggrandizing rhetoric, rendering the whole exercise useless.

Interesting that that old NYCB entry should come up today, because I just got to the part of The System of the World with the remote-control garotte. (Well, more or less.)

1. The 204 HTTP status code, "No Content". In addition to denoting an extremely boring website, you can return this and the client will stay where it was instead of expecting a new page from you. This is great for control panel type pages, in conjunction with some Javascript on the client side to mantain displayed state (I know, I've been seduced by the Javascript side, or perhaps just the client side).
2. Python open source job opening writing test automation software.

Then I realized that you could use the exact same mnemonic to remember this as well. In spring you go on Daylight Savings Time, and in the fall you go off it. Spring on, fall off. Pretty slick, no? Too bad they already gave out the Nobels this year.

The other thing you can do is live in the UK, where the name of the Daylight Savings timezone is British Summer Time.

Since I'm not afraid of power outages due to ext3, it's now sitting on my computer room floor enjoying an undeserved retirement. That's the story of the deadbeat UPS.

Of course, all they did yesterday was give fancy names to post offices. Maybe we should be grateful for that, actually.

Does anyone have any suggestions of how to get more getCosmos()-like behavior without actually cheating? All I can think of is to scrape del.icio.us instead, but that puts me in a del.icio.us ghetto. I may end up building my own private mini-Technorati.

The system keeps two sets of books, and you can get it to display tallies from the fake set instead of the set that reflects people's actual votes. Paper verification ballots are useless, because the software knows to use the real data for spot checks and only uses the fake data to report aggregate results. This is really bad. It undermines the integrity of the election system.

Something like this has long been suspected, and it is a general problem with electronic voting and vote counting systems, but it is no longer a hypothetical problem. The attack actually exists and has been demonstrated. (That link has a very detailed article about the problem which I recommend you read.) The system has been in place for four years in 30 states. It's possible it has already enabled election fraud.

As in accounting, keeping two sets of books is not something you do unless you want to cheat or are being pressured to cheat. It's not a bug--it's a whole other system designed for cheating, hidden inside the system the states were sold. The authors of that article pin the blame on a specific person, who had means, motive, and opportunity. But who specifically did it is not as important as the reaction to the discovery of the problem, which has been just awful.

Let's suppose I did something like this at the company where I work, and play out the scenario that would happen once someone found out. Obviously I would be immediately fired. My now-ex-employer would inform all our clients about the problem, and they in turn would demand that 'we' fix the problem immediately. Most of the bigger clients would ask that their sites be shut down until the problem was fixed. Everyone would take the problem very seriously.

While it would not be in my now-ex-company's interest to have information about the problem made public (at least until it had been fixed), it would also not be in their interest to let the problem remain unfixed. It would also not be in the client's interest to ignore the problem. If anyone ever exploited this hypothetical problem it would cause enormous damage to the client, and if we didn't fix it quick they would fire 'us', quickly.

Now let's exit that hypothetical and see what happened in the real world. As far as I can glean from that article, (it's a little vague), the person they blame doesn't work at Diebold anymore. I don't know if this is because he was fired, or because he got sent to prison for (unrelated but motive-providing) embezzlement.

But that is all that's happened. Nobody seems interested in fixing the problem, and according to the people who know how to exploit the problem, a lot of the time they can't get the interested parties to even watch a demonstration.

What's wrong with this picture? Here's my guess. Note that I'm going to explain this situation without claiming that any of the parties currently involved want to rig elections, which is certainly a possiblity but which explaining things that way tends to get you dismissed as a conspiracy theorist. Anyway, the problem exists and is exploitable whether or not I'm a conspiracy theorist.

Diebold is (in my explanation) trying to keep the counties and states from demanding action because any effective action would result in Diebold losing the contract--remember, this back door has been in the software for years. It's already way past "fix this now or else" territory and into "we don't trust you and we never can" territory--if the state wanted to take action.

The state officials don't want to take action because they don't want to admit that such a huge problem could happen on their watch. They know that if this knowledge becomes widespread the public will blame them (with some justification). A lot of them also have a lot of political capital invested in "information superhighway" type electronic voting rhetoric, and the states have put a lot of money (by state budget standards) into these fancy systems. It would cost a lot of time and money to switch. And for what? So you can say "Yes, this election was secure."? Well, you can say that no matter what, and save the time and money.

There's also no upside to solving the problem for the state officials. They are not directly damaged by bad elections so long as no one finds out they were bad (unless their opponent rigs the Secretary of State election against them, which would be ironic but also difficult). The only reason they should try to solve this problem is that their job is to not let this problem happen. California's secretary of state, Kevin Shelley, is the only official I know of who is doing his job in this respect.

If there were consequences for ignoring the problem, there would be incentive for the state officials to fix it. We wouldn't have to depend on government officials wanting to do the right thing. Right now there are no consequences because for the most part the ultimate bosses of everyone in this scenario (the citizens of the states) don't know about the problem yet. If the state officials watch a demonstration of the problem, they destroy any plausible deniability they might have had, and they create a news hook for media investigations into the problem. That's why they don't want to even look at it.

I am posting this entry to do my part to get rid of the plausible deniability. It's very important that the states secure the application as best they can for the coming election, and switch to a more secure system as soon as possible.

The del.icio.us plugin is in NewsBruiser, but I think it's stretching the limits of what should be in NewsBruiser. I can't see that many other people really wanting to echo certain of their weblog entries to their del.icio.us accounts. Am I wrong?

I know what you're thinking: "Great, another hacker trying to use the hammer of technology to hit the non-nail of law. Also, he has poor personal hygiene!" Well, hear me out before you judge, hoo-mon. My personal hygiene is actually quite good, and look at this hypothetical complaint constructed by the EFF to demonstrate how awful and exploitable is Orrin Hatch's INDUCE (now IICA) bill[0]. I'm sure my prattling to Seth had nothing to do with this, but it is the perfect example of what I was talking about.

Note that in a twist rarely seen in the technical realm, sometimes the people writing the law are actually using the law as cover to enable some new attack. In that case it's even more important for outside reviewers to find and expose the backdoor in the law. Also, all laws should be written in my Leibnitzian Python wonder-language that contains no ambiguity.

[0] Leonard's Analogue to Leonard's Analog[sic] to Brooks' Law: "Renaming a bad law makes it worse."

Anyway, on one of the trips coming back from Bakersfield, what should be beamed down from the XM satellite but a song called "Cool Off", by New Orleans ensemble Galactic. It was nice, funky, jazzy. But it just went on and on. For minutes! And there were no words.

I don't know how much you know about me, random weblog reader, but I've got to have words with my music. When there are no words, I get cranky and I skip the track or change the station, or I start just making up words and belting them out. In this case the beat was groovin', so I made up words and had impromptu karaoke.

I was in me a boat crash
I crashed into the shore
Flew out of the cockpit
I wasn't drivin' that boat no more

Chorus:
You better cool off, cool off daddy,
Cool off daddy-O.
Cool off, cool off daddy,
Cool off daddy-O.

I was in me a chess match
Moved knight to queen's rook three
My opponent got angry
Started beatin' the tar out of me
He said:

Chorus:
"My name is Kulov, Kulov, daddy,
Kulov, daddy-O.
Kulov, world chess champion,
Kulov, daddy-O."

I was down me in Bolivia
Said some things made some people upset
That's when they pushed me against the wall
Offered me a last cigarette
I said:

Chorus:
You better cool off, &c.

Eventually the guy on the recording started thanking everyone for coming out tonight, so I assume the song was the last track on a live album and its purpose was to "cool off" the audience after a hot jam session. My words are still king, though.

Confidential to Jake: let me know if you're interested in laying down some jazz mix for my goofy lyrics, and I'll send you a vocal track. Think of it as like "What's A Godly Person?", except I know you're doing it.

This is, incidentally, the best thing about Alton Brown's style of cooking pedagogy; he shows you the makeup of dishes, how to analyze them, and the connections between them. One of the best examples of this is that in some episode or other of Good Eats he points out that a cheesecake is structurally a custard, not a cake. It needs to be cooked in a hot bath, not just baked in a pan just because it has 'cake' in its name.

Anyway, for some reason I didn't buy that book! It cost $10 and I guess I decided it wasn't worth it. Later I changed my mind, but the last time I went to that bookstore it was gone.

Today I idly searched Bookfinder for "cooking patterns" and I found it! I don't remember the (long, unwieldy) title and I can't search for it anymore because I bought it and there are no other listed copies, but by a fortunate confluence of minds the author put "patterns" in the title so I was able to find it. And at about the same price it was being offered at the used bookstore! The $3.00 extra I'm paying to have it shipped to me I will chalk up as a "lesson learned" surcharge. The relevant lesson, of course, is: never take chances! Buy EVERY BOOK!

If I ever write a cookbook (which I might) it's going to be a synthesis of this cookbook I've got coming to me in the mail and Alton Brown's I'm Just Here For The Food. It'll be organized around the principles of recipe schemas and recipe transformations based on ingredient or technique. If you can make x, you can make y by applying a transformation. You can start slow, gradually build a repertoire you're comfortable with, but always keep experimenting within the bounds of the known if you don't want to learn a whole new skill to make something different.

I think cooking should be more like quantum chemistry and less like high school chemistry. There's always going to be a subjective element, but I think it can be factored out and that you can learn even from reading about transformations that involve stuff you don't like.

The two engines give pretty similar pictures, except that Yahoo says public domain usage and the "Attribution" license are much less popular. Note the following:

• Every Attribution license is more popular than every non-Attribution license.
• The non-Attribution licenses are popular in approximately the same order as the Attribution licenses. The exception is the ingeniously perverse "Sharealike" license, which is sort of like a viral public domain.
• Yahoo and Google agree that the the top 4 licenses have about 84% market share, and that the top 5 have over 90% market share.

I guess I knew about metal combustion before, because I'd seen the pretty spectrographic colors metal dust makes when thrown into a fire. And it's not surprising that for metals with nearly-empty outer electron shells, a fine line would divide oxidation from combustion. But never before had I envisioned metals catching fire on their own and wreaking havoc, or conceived of high-tech fires deemed worthy of their own fire class (D) and requiring special fire extinguishers to appease. Even though your aluminum foil isn't going to catch on fire, the idea has a certain cutting-edge cyberpunk allure.

Sensors also available in "boring box" edition. Maybe it has a compartment for you to hide your house keys in, or something to make up for the fact that it doesn't look like a frog.

Site does some stupid Javascript thing so I can't be bothered to find better links. Produit de France.

Idea follows. Compute a one-way hash of your email address. For instance, "8zhkqqkZ.A9tg" is a crypted version of "leonardr@segfault.org". "b83aaa5e5c8129987c8c54f2974e84d3" is the MD5 sum of "leonardr@segfault.org". You get the picture. Then give out "b83aaa5e5c8129987c8c54f2974e84d3-whatever@mysecretemail.com" to someone from whom you want to hide your true email address. When you want to see if they sent you mail, go to mysecretemail.com and say "Hey, I'm leonardr@segfault.org". mysecretemail.com thinks "Oh, that's good old b83aaa5e5c8129987c8c54f2974e84d3.*@mysecretemail.com," and forwards you all the accumulated mail.

It's cool and secret-agent-ish, but like I said I can't think of a real use for it. Also you could break it if you had a way of generating a specific hash collision, which you kind of do.

PS: more useful, less cool variant: leonardr-at-segfault.org-whatever@mymyriademails.com. This lets you get get infinitely many forwarding accounts for one email address.

All that was seemingly a big lead-up to the fact that I've had to start putting up "misc" directories to contain pictures that don't correspond to a "big event, better take the camera" thing. Here they are for February and March. These sets include hilarious moments like At least they didn't park in the handicapped space.

Bonus for Joe: this Ad campaign for New Zealand found near the Apple store which opens up new fields in tautological advertising. "Don't even come," it says. "We won't let you in."

Bonus bonus for everone: for some reason they're ripping up all the greenery around the building where I work. I assume they know what they're doing. Anyway, I call this picture set The Scouring of the Shire, and it should satisfy those of you who were hoping Return of the King would be even longer.

"What animal are you?"

"Human."

[GLARE]

It turns out I am actually a sheep. I was thinking it would be nice to vary the Chinese lunar calendar to give other organisms a chance. We would have cool years like the year of the elephant, year of the saber-tooth tiger, year of the mola mola, year of the E. coli, etc. Just another idea from Leonard Labs, where we're always working orthogonal to your best interests.

"For service to an existing building, press 1."

What should have come next:

For service to a building that has yet to be built, press 2. For service to a demolished building press 3. For service to a hypothetical building, or to the concept of buildings in general, press 4.

But I'll tell you what does exist is solar cookers, a real New Cooking Technology for the Third World. They are trivial to build, and you can use them to cook and to purify water. In the ultimate mockery of Mother Nature, there is even a solar refrigerator. Am I the only one who thinks there's a vast untapped market for science fiction about this kind of thing? Is this what's in all those Bruce Sterling books I've never read?

Backyard Artillery also has other weapons of varying degrees of lethality, including suction cup crossbows, childhood favorite burp guns[0], and labor-saving hand-held bubble machines.

If you're into the heavier stuff you can build or buy ancient siege machinery and revive the gentlemanly art of laying siege. (Siege defenses, hunger, privation, reduction to last extremity, traitor crushed by shields all sold separately.) Provided history of siege machinery includes amusing yet frustrating detail:

Unfortuantely, we now know that most of [Ralph Payne Gallwey 's] sources are fiction. He even invented fake reference numbers for them at the Bibliothec National in Paris, France.

All these links derived from cl links.

[0] Susanna, I'm sorry I shot at you with the burp gun at our grandmother's house when you had no way to defend yourself.

I remember when Magic Markers were invented; the first ones my mother brought home were about 1956. They consisted of a small, chunky glass ampule filled with noxious ink and a thick felt wick. I tried them out on the living room walls and found them to be a satisfactory art medium. However, post-creation I abandoned them to the floor, where my father's big work boots trod upon one of them. Bye bye Magic Marker.

I can either stay at home tomorrow and wait, or redirect the delivery to my work address. In either event I will not actually be able to use the DSL until Wednesday night, the day before I leave for Utah. Oh well. Not a big deal. Note my lack of rage; it's all part of the new kinder, gentler, less full of rage image I've been told to cultivate for my gubernatorial run.

The old electric is still available; I'm going to give it to my grandmother to use in a garage sale, so in the unlikely event that anyone wants it, let me know and I'll rescue it.

1. Kick-spinning tripod of death
2. Technology enabling "Post-Feudal Ninja Industrialists" to be the first movie filmed in NINJA-VISION!
3. Neighbor to the Temple Mount

I was disappointed, because even though British and French taxpayers will no longer bleed billions of dollars per year, the rest of us lost a great shorthand for fallacious investments :).

The reason your precarious picture doesn't look precarious is perhaps a lack of context: you know that the drop is steep, but there is no non-steep stuff in the shot. If you had got the horizon in there it might have looked more precipitous.

Incidentally this is why everyone who stands at the bottom of a building and takes a shot up at it is disappointed with the lack of looming in the final image. You need some context.

Someone get this man a cable talk show!

I have a feeling that the cause is that your camera has a very wide depth of field. If the stuff beyond the drop were out of focus, and the precipice itself in focus, then the precarioutiveness might thus be captured.

We're simple folk 'round these parts. We just point and shoot. Sometimes we zoom.

I got my AS&S catalog in the mail yesterday, and I'm looking forward to another romp through the cheap but useless. I recommend signing up for the catalog; it's a lot of fun, and they don't seem to sell your address to other catalog companies.

Pete Peterson II is also in the enviable position of having the actual AS&S store "just down from the street" from where he lives. I used to imagine the AS&S store as an enormous warehouse in the middle of an Illinois cornfield, maybe next door to a pesticide reseller with a corrugated metal roof, but Pete says "it's just a medium sized store inside a strip mall that is wall to wall packed with stuff."

My dream useless 'convergence' machine would be a combination Geiger counter/GPS system/ham radio. Fearmongering use case used to sell this device: "Help, I'm trapped near an old Soviet test site in Kazakhstan!"

Today my car thought "That Leonard thinks he's pretty smart. He's fixed the gas tank cover latch hook, but now it's time for LEVEL TWO!" I went to the dentist this morning, missing the shuttle, so once again I drove into work. As I drove out of the parking lot this evening I heard a crunch. One of the little plastic connectors that keeps the front spoiler[0] attached to the car had busted off, and my spoiler was dragging on the ground. I retraced my steps, found the connector, and drove gingerly home where my tools[1] are. It turns out I didn't need my tools because the connector is an ingenious two-part assembly, and once you figure this out you can just reattach the spoiler to the car with your bare hands. Success again!

Tune in tomorrow, when my hubcaps fall off and must be replaced with dinner plates.

[0] This is the second time this spoiler thing has flaked out on me. The first time this happened I didn't realize it until the connector was long gone, so I had to take my car into the shop, and that's what the guy called the thing in the front of the car. I don't think "spoiler" is the right word for it, but that's what he called it. I'm just sayin'.

[1] These are computer- and minor-home-improvement-geek tools, not car tools.

As with so many others, my dream job would be something for the United Nations Godzilla Countermeasures Center. My dream job that actually exists (again, as with many others) would be doing cool things for Google. (Today I actually thought of a cool thing to run against Google, of which possibly more anon, but it won't work without angering Google because it won't work within the rather limited Google API).

From clickolinko: Swappington's, a cultural-artifact-trading site with a minimalist interface, an interesting (and, I fear, exploitable) economics, and a goofy animated GIF demonstration. (Incidentally, if you decide to get an account there, list me [leonardr] as your referral, since referrals are the only action I can find that create money.)

From Boingboing: the passive-aggressive approach to protesting workplace patent requirements.

• My Hiptop broke last Sunday.
• On Monday night I called T-Mobile to get a replacement shipped to me.
• I was told it would be shipped in 24 to 48 hours. I asked to have it shipped next day air, so that I'd get it before leaving.
• I was told to return my nonfunctional Hiptop (postage paid) within seven days or face stiff penalties.

At this point I began concocting elaborate disaster scenarios involving my vacation to Bakersfield and delays in their sending the Hiptop. These worried me until last night, when I called worriedly to make sure they'd shipped it and I'd get it on Friday.

• They hadn't shipped it yet.
• I couldn't cancel the order.
• I couldn't get the address on the order changed, though the operator promised to try to get them to make an exception for me.
• I would get a notice via email when my package had shipped.

My worst fears--confirmed! I sank into despondency, forseeing delaying my trip south for days of waiting at home for the postman for a package which never arrived, bureaucratic nightmares at the post office, etc. etc. I held out hope for the possibility of one scenario, so unlikely that I felt foolish even considering it.

• Today the Hiptop arrived.
• I have yet to receive email saying that my package has shipped.

They had been entirely accurate in their promises the first time I'd talked to them, and wrong in every detail when I called to follow up![0] Amazing! It's a Christmas miracle, adjusted for inflation!

So, I have a working Hiptop now, and I'm happy. A couple times in the past 12 hours I began NYCB screeds against T-Mobile's customer service but deleted said screeds as petty, vindictive, and uninteresting. Determined to learn from my earlier rhetorical excesses, I was working on a calm, reasoned indictment of the bastards when the point was rendered moot.

[0] It's conceivable that my package could have been shoved out the door the minute I got off the phone, but it was sent through USPS, which shuts down at 5, ne c'est pas?

We joked about it, Kevin and I, but joking about it doesn't make the problem go away. Today whilst browsing Clifford Pickover's new blog (man, it just gets worse, don't it?) I saw a link to this New York Times story (and again!) about Internet-connected videocameras in coffins. I haven't read the article due to the whole requires-registration thing, but the angle seems to be shock value, whereas I think guilt is the big potential money-maker here. The ancient art of Google-fu (and, I suppose, the even more ancient art of looking at the URL), reveals the article to be about a Dutch made-for-TV movie, so for now at least the idea remains safely in the realm of scathing social commentary made-for-TV movies.

I'd post a random picture taken from the Hiptop, but the web browser doesn't seem to support file upload. (Finally, complaining, my natural medium!) More Hiptop news as it develops (but probably typed on the big computer).

(Later) Here it is.

It's all pretty useless, of course, but my cube at work is full of useless stuff that isn't connected to the computer, so why not (for example) make my Godzilla action figure stomp around whenever someone assigns me an issue?

Mike's weblogs provide much-needed competition for Danger Info, which has become a bit complacent recently. Actually, I just made that up.

My main problem with Snow Crash is its manifestation of what I once called the Fundamental Cyberpunk Error: the FreeCiv-ish idea that civilization consists of a Fisher-Price hammer-and-peg playset with a bunch of discrete technologies and social constructs sticking up, and that you can tap on one with a hammer and push it down without it having any effect on any of the others. "Of course there will be sports in the future... [tap, tap, tap] DEATH SPORTS!" And you have Rollerball.

Example: setting completely aside the usability problems of virtual reality, how can a fully immersive high-bandwidth world-wide virtual reality universe continue to exist in a world without the rule of law or the sanctity of contract? Who mantains the servers (or other electronics)? Who manufactures the servers (or other electronics)? Who mines the raw materials and how do the raw materials get to the fabrication plant without being stolen by bandits? Who grows enough food to feed all these people working on assembly lines instead of hunting and gathering? Who maintains the microwave stations and transatlantic fiber optic cable, or launches new sattelites into space to replace broken ones? How does each party to this operation afford the cost of the private army required to avoid getting ripped off or blown up by rivals? It's to solve these problems that people form states[0], but once you've tapped down the little "State" peg with your little cyberpunk hammer you don't have that option.

Anyway. My point here is not to carp on particular problems, but to discuss this sort of inconsistency in general. I carp not on particular problems because Sumana convinced me that microlevel inconsistencies can happen in a cultural artifact even if the long-term cultural shift in that artifact's universe is in a particular direction ("The market can stay irrational longer than you can stay solvent."), so it's not prima facie evidence of poor craftsmanship to include such inconsistencies. I find myself much more favorably disposed towards Snow Crash, and I think it does the book more justice, when it's regarded as a snapshot of a civilization seventy-five years into a four-hundred-year decline into tribalism and anarchy rather than (as I regarded it until recently) as a picture of a civilization already completely collapsed into tribalism and anarchy.

The thing is that this exact same sort of inconsistency happens in cultural artifacts which are generally agreed to be awful: for instance, Mad Max Beyond Thunderdome, which nobody except Jake thinks is great.[1] (Paraphrase: Jake: "all my friends think mad max beyond thunderdome is a terrible movie." Leonard: "Has it ever occured to you that there might be an empirical basis for such a belief?" Jake: "that is impossible.") It happens in Signal to Noise, a book which may well have merit (though I didn't like it) but in which all the characters act as though the American government still has power, even though voters stupidly passed a poorly-written Constitutional amendment which has as a trivial side effect the total emasculation of government (perhaps the case that will establish this is still awaiting certiorari).

I am ill-disposed towards inconsistency in general (I once rewrote a song because it contained an inaccurate statement about evolution), but I think I can now distinguish the good kind of inconsistency (forms of archaic rituals from the modern era preserved through inertia and other microlevel eddies flowing against the macrolevel, enhancing the richness of a story) from the bad kind (consequences of authorial decisions not properly thought out, causing gaping plot holes and annoying me). It comes down, I think, to judging an inconsistency against the gestalt of the book. For instance, A Canticle for Leibowitz has some of these inconsistencies, but they seem to me like the good kind, and I really like A Canticle for Leibowitz. I haven't read The Postman, but from what I've read about it it seems the very embodiment of the good kind of inconsistency, and I've a feeling I would like it as well (though apparently the movie is horrible).

[0] I know, Locke was wrong, people don't 'decide' to form states to solve particular problems--but once you have states, it's things like the rule of law that distinguish states whose citizens can create things like the Internet from states whose citizens can't. And any cyberpunk-esque mutual defense venture created to get around the lack of a state is in fact a Lockean state of the sort that people don't decide to form. This is actually one of the premises of Snow Crash, which means I'm arguing in circles--wheels within wheels, Jeeves. My point is that to mantain the civilizational infrastructure neccessary for an reliable virtual reality Internet, your de facto state must have power and agreements with other de facto states consumnate with the power of and mutual agreements between today's nation-states, so why not put the microstate idea in another book so as to do it justice?

[1] In British radio programmes this is known as "the sort of statement that gets us letters".

Mechagodzilla is really the coolest Godzilla monster ever, apart from the G man himself. He's the poster boy for government waste. "Gentlemen, once again an immense kaiju threatens our shores. The only solution is to build a full-scale replica of Godzilla out of dully-gleaming, plastic-like metal."

OK, my huge download of the tarball of RPMS containing the Tigris runtime environment is done. Back at it.