The Only Winning Move

IBM has apparently developed cat brain software. Don’t rush out to buy it, though. You need 147,456 processors to run it - not something that you can fire up the Python interpreter on your average machine and expect a return value from. I think the point is just that this is really motherfucking impressive. And heck, I’m impressed.

My comments.

First - the obligatory political observation about how it didn’t take the NSF to bring us this apparently useless and profit-less science project. Private companies can and in fact frequently do fund research for the sake of research on huge scales like this one. So the left can stop with that “there’ll be no pure research AT ALL without the government!” straw man any time.

Second - in response to all the snide comments on the web about how IBM is running a cat’s brain because cats aren’t actually all that smart compared with dogs, I think you’re all still missing the point of this pet raising thing. Pets are companions, not calculators. Even if it’s true that dogs are smarter than cats, I’m not sure how that bears on my decision about what kind of an animal to feed and spend time with. Isaac Newton, by all accounts, was a really smart guy and a first-class whiny asshole who had no friends. It turns out you can be smart and not all that much fun to be around. Which is what dogs are. They’re useful as tools for things like sniffing out bombs and criminals and participating in rescue missons, but spending time with something that slavishly devoted to you is just not interesting. Which is a polite way of saying that dogs may be more intelligent than cats as animals go, but people who like dogs aren’t as smart and don’t have as much character as people who like cats, and that’s the real issue, kids. So try another line, will ya? Or take up the courage of your convictions and raise pigs instead, since they’re definitely smarter than dogs. I hang out with cats because they’re fun to hang out with, and that’s really all that’s involved.

Third - cut the crap about how this is scary. It’s not. It’s not even a whole cat brain - just some subset of a cat brain. It’s certainly not a human brain, and even as a cat’s brain it runs 100 times slower than the real deal. You can’t stuff it in a robot aibyou (note the ‘y’ - that’s “pet cat” in Japanese, rather than pet dog) and have it chase string! And as for the people who say that if Moore’s Law holds the same number of processors will be able to handle a human brain by 2019, well, you’re making a lot of assumptions. First, you’re assuming that there’s any evidence at all that this giant simulation of a cat’s brain is really behaving like an actual cat, and I haven’t seen any provided. Second, you’re assuming that Moore’s Law will continue to hold, which, since it doesn’t even hold now, seems unlikely for the future. Third, you’re assuming that there’s anything at all threatening about a human brain living inside a giant computer complex, which there isn’t. Fourth, you’re assuming that machine intelligence must be functionally equivalent to human intelligence to count as intelligence, which is just silly. I forget who said it, but there’s a money quote from someone saying “The question of whether machines can think is about as interesting as whether submarines can swim.” Which is to say, it’s all a matter of semantics. Machines already do think in non-trivial ways, and what it is exactly that you’re (irrationally) afraid of will be here long before we’ve made a convincing machine implementation of a human. You COULD, but you WOULDN’T, build a computer out of hydraulics, just like you COULD, but you WOULDN’T, implement machine intelligence by simply rebuilding a human brain, only this time with silicon! There are much more efficient ways to achieve machine thinking.

Well, the future is here it seems. But then, it always was, wasn’t it?

The ongoing saga of trying out .NET on Mac OS X via Mono. Remember that libiconv thing, where installing Mono briefly and mysteriously broke Mutt? Well, it turns out it breaks MonoDevelop too! More proof that the Mono Development Team isn’t exactly Mac-friendly. No - that’s NOT an accusation of sneakiness! The Mono Team does great work, and I’m grateful that Mono exists at all. That it runs as well as it does is a testament to their skill and dedication. No, I’m just taking this as more evidence that Mono doesn’t get tested as thoroughly on the Mac platform as it does on Linux and Windows - owing, I speculate, to the preferences of the development team. Fair enough.

In any case, the issue is that you (meaning some subset of us Mac users MAY) have to update libiconv to get Mono to work in the first place. Interestingly, though, the official beta release of MonoDevelop seems to depend on the same outdated version of libiconv that crashed Mono for me in the first place (at least, that’s what the error message I got implied…). Then, in a final and somewhat bizarre twist, it turns out that just grabbing MonoDevelop from a link on Miguel de Icaza’s blog directly (download link here: looks like it’s an early one that M J Hutchinson put up and never took back down) resolves the whole issue. Which is strange, because it makes it look like MonoDevelop for the Mac first used the new improved libiconv, then switched back to the old one inexplicably. Maybe in response to bug reports? But then, why does Mono itself continue to rely on the new version?

Clearly I’m missing something here, but I’m not complaining. I’m happy to have both Mono and MonoDevelop working and am looking forward to trying out the ASP.NET MVC here on the Mac. Thanks again to the Mono team for bringing .NET to the Mac platform - it’s much appreciated!

Well, wouldn’t you know it. My two big log pages for “how-to-install-Mac-unfriendly-shit-on-Leopard” turn out to be incompatible! I KNEW that installing Mono would break SOMETHING, and that something turns out to be Mutt.

The problem is that the library that Mutt calls to do its unicode conversions is libiconv.2.dylib, which was of course one of those things that had to be upgraded to get Mono to work. Identifying libiconv as the problem wasn’t the hard part; what had me gussing was that trying to reinstall Mutt via MacPorts mysteriously didn’t work.

It’s not a black mark against MacPorts. It turns out that I have been using a devel version of Mutt, which means I should be prepared for things to break while they iron out the kinks. So recompiling it from source with the new version solves everything since the new version doesn’t choke on libiconv.

OK, that’s not quite true. The Mutt team seems to have removed a bunch of features, so I had to comment out a lot in my config file to get it to load without a spew of pointless error messages. And it does this weird thing now where it doesn’t play all the headers back to me as I’m composing an email. Somehow the signals haven’t quite been reconnected in some parts. I don’t know if that’s my configuration or some stuff that’s still “under construction.” I’m guessing it’s my configuration somehow, but I’m sort of unwilling to start from scratch on that with an official config file. Maybe this weekend.

Anyway, it’s good to have the world’s best email client back up and running. I’m pleased to have Mono, but life is seriously hard for me without Mutt!

OK, I’ve had some trouble with browsers quitting unexpectedly from time to time, but this has got to be a first: twice today Opera has not shut down when I’ve asked it to. Doesn’t matter whehter I use the mouse or just hit apple-Q, it chugs blithely along. Very disturbing. Force quit still works, of course.

I am a big fan of the Mutt email client. Unfortunately, I also really like my iMac, and getting Mutt to work on it is always a headache. So I usually just retreat into my shell and put up with Mail.app. It’s not bad, after all. But it’s also not Mutt.

I have a Dell running Gentoo at Alexis’ house, and working on that recently I was reminded what a sine qua non Mutt actually is. So this morning I sucked it up, rolled up my sleeves, and got to work.

First, of course, download Mutt. I save it to desktop, so the next steps are obviously:


cd ~/Desktop
tar xvzf mutt-1.5.18.tar.gz
cd mutt-1.5.18


Then you gotta configure it, and this is where the trouble always shows up. Mutt (as of 1.5) comes with smtp support built in - which is a minor violation of its design philosophy, but was much appreciated back when I was still unaware of the magic that is Getmail. Even with Getmail in the background, I still compile with smtp support … just in case. These instructions were borrowed from linsec.ca:


./configure --prefix=/sw --with-curses --with-regex --enable-locales-fix \
--enable-pop --enable-imap --enable-smtp --with-sasl=/sw --enable-hcache --with-ssl --mandir=/sw/share/man


The trouble for me always comes with the


--enable-hcache


flag. This requires a nice database manager in the background - apparently either QDDBM, GDBM, or Berkeley DB4. So, I installed Berkeley DB4: Berkeley DB 4.7.25.tar.gz.


cd ~/Desktop
tar xvzf db-4.7.25.tar.gz
cd db-4.7.25.tar.gz/build_unix
../dist/configure
make
sudo make install


Which installs Berkeley DB4 in /usr/local/BerkeleyDB.4.7/

And that, ladies and gentlemen, is the problem. Mutt isn’t expecting to find it there, and I then generally spend a couple of hours pulling my hair out trying to figure out why this thing that I clearly installed can’t be seen by Mutt. I’m not completely stupid, you see: I do know about things like the CPPFLAGS and LDFLAGS options to configure, which are meant to tell the compiler where to find libraries it can’t find in the usual/expected places. For some reason I don’t understand, these have never been any help compiling Mutt. But it seems there is a command line flag that tells Mutt exactly where BerkeleyDB4 is if you installed it. Here’s the list again; see if you can spot the difference:


./configure --prefix=/sw --with-curses --with-regex --enable-locales-fix \
--enable-pop --enable-imap --enable-smtp --with-sasl=/sw --enable-hcache --with-bdb=/usr/local/BerkeleyDB.4.7 --with-ssl --mandir=/sw/share/man


Yes, apparently –with-bdb=/prefix is there for the express purpose of telling Mutt where to find the database. All this time and I never knew… Well, no one but myself to blame - it’s right there in the configure script if you bother to read it.

At this point Mutt compiles without complaint. You’ll notice that the guy at linsec.ca (Adam someone - whose name I should probably be able to reconstruct but can’t) asks everything to install at /sw, which is where things installed by Fink tend to go on Mac OS X. I put it here too, having no better ideas about where it should go - and if you choose to then you’ll have to make it visible to the main system in some way. This means you either add /sw to your PATH variable (probably via ~/.profile), or you make a symbolic link to some folder such as /usr/bin where your system is already expecting command-line programs to be:


cd /usr/bin
ln -s /sw/bin/mutt mutt


Now you should be able to type mutt on the commandline from any directory in your system and be in business.

The next step is the dreaded configuration of sendmail. Unless I’m very much mistaken, sendmail is already on your Leopard system (it was already there for me - though I’m blonde, so it might be there because I installed it at some point in the mist-shrouded past of a month ago and subsequently forgot), so you don’t have to install it. But you do have to get it to do what you want, and as with anything this can be a bit tricky…

…which is why I don’t bother. There’s also this bit of black magic called putmail.py - as the name implies, a Python MTA. Actually a fairly minimalist one: it’s tailored to people like me who need to send mail, but not do anything fancy - just username, password, send, thank you very much. And that’s what it does for smtp. Installing it and setting it up is the easiest thing in the world - just read the helpful man page and it’ll take you 5min. tops. The config file goes in ~/.putmail/putmailrc. Here’s mine:


[config]
server = smtp.server.name
email = my_email_username@domain.name
username = my_email_username
password = my_password
tls = yes


(The last line is necessary to get it to go through IU.)

Now we have the ability to send mail and read it, but we’ve still gotta actually get some mail to read.

The standard thing to do here is use fetchmail - but I feel the same way about that I feel about sendmail: too much trouble so long as there’s Getmail in the world. The link goes to the latest platform-neutral version unixy version. Getmail has the same basic advantage that putmail.py does: it’s written in Python, so it installs easily and works out of the box whereever a cool enough (read: >2.3 at the time of writing) version of Python is installed:


tar xvzf getmail-4.8.4.tar.gz
cd getmail-4.8.4
python setup.py build
python setup.py install


DONE! Well, except for the configuration. Configuration files go in ~/.getmail/getmailrc. Like with putmail.py, it’s all so simple there’s little point in walking anyone through it - but for reference here’s my file for talking to my Gmail account:


[retriever]
type = SimpleIMAPSSLRetriever
server = imap.gmail.com
username = my_username
password = my_password
mailboxes = (”INBOX”, )

[options]
read_all = false


A couple of things to point out. First - this tells Getmail to store my mail in mbox format, which is pretty standard UNIX, but may not be right for everyone (basically - it’s a giant text file with all your mail in the same file). And I’ve also told it to call this mbox file “inbox” and put it in a folder called “Mail” in my home directory. I mention this because it’s important to tell Mutt where to look for mail. Finally, I’ve set read_all to false to make sure it only gets those messages it’s not gotten before. Fail to do this, and you’ll end up with tons of duplicates in your inbox. Another way around this problem is to do delete = true, which deletes messages from the server as soon as it retrieves them. I go in and set it that way about once a week to do a quick mailbox cleanout - but otherwise it’s nice to leave about a week’s worth of messages on the server so that I can run mutt on several computers (laptop, iMac and Gentoo machine) without having to cart my mail between them. (More accurately, you have to toggle this on and off - because when set to false it only retrieves mail that’s never been read in any session, so it is a minor headache. Probably there is a good solution to this, but my version is just to delete the inbox mbox and redownload each time - poor man’s version control. When you’re keeping your mail to within a week, it’s not inconvenient.)

You do this - where else? - in your muttrc file - which it’s customary to hide in plain sight in your home directory, like so: ~/.muttrc. The best thing to do is copy the one they give you with your Mutt distribution - which has all the options typed in and commented out. You uncomment the ones you want to use and change the values to taste.


cd ~/Desktop/mutt-1.5.18/doc
cp Muttrc ~/.muttrc


One can, of course, spend days fine-tuning Mutt to taste (which is the whole point of it, really - a mail client that does what YOU want for once), but the bare minimum to get it running is this:


set mbox="~/Mail/inbox"
set mbox_type=mbox
set sendmail="/usr/local/bin/putmail.py"
set spoolfile="~/Mail/inbox"


And now it does the basics! If at this point in your life you type getmail on the command line, wait for it to finish and then type mutt, you’ll be reading your mail in the world’s coolest mail client. Naturally there’s no point in dealing with Mutt if “the basics” is all you want - but from this point it’s all a matter of what you like. So - go through your muttrc file and/or read the excellent Woodnotes Guide to the Mutt Email Client to get you started and have fun with it!

I’ve seen two iconic “hacksploitation” flicks in as many weeks: Hackers and WarGames. Whether they deserve to be or not, these two somehow managed to end up the representatives of the genre for the 1990s and 1980s, respectively. I will now load my taste module and tell you what I think of them.

Alright, I don’t really ask Roger Ebert’s opinion on everything, I just wanted to point out that this is one of those he got more or less right. I would’ve given Hackers a lower rating than 3 stars myself, but Ebert redeems himself yet again another time already by recognizing WarGames for the classic it is. It gets a well-deserved 4 stars, and more importantly, he “gets” that it’s superior to its 1990s counterpart.

Why?

That’s actually something of an interesting question. There’s a lesson here for how to make movies - indeed, how to tell stories in general - that needs pointing out. When I was brainstorming about this myself, the answer I kept coming back to was that WarGames is just “more real” than Hackers. But obviously that’s absurd. Hackers, for all its many flaws, actually has something like a realistic plot. Trying to show off to his hacker friends, a kid hacks an oil company and does a half-assed job of it, getting caught. Normally the company would let it go, but unfortunately the security expert who catches him has been indulging in some salami slicing, and the kid just happened to have downloaded an incriminating file. So rather than take the fall himself, the security expert decides to frame the kid, and the rest of the story is, in addition to being about a rivalry between two hackers, largely about trying to put the blame back where it belongs. I mean, when you really step back and think about it, this is not only “realistic,” it’s probably actually happened not once but several times. Some security experts estimate that salami slicing happens all the goddamn time without ever coming to anyone’s attention. And what criminal won’t frame an amateur if he thinks doing so will get him off the hook? Hacking rivalries, for their part, are real and common. So what’s not to like? Contrast this with WarGames - where we have at the center of the plot an AI so advanced that it isn’t even on the horizon now, let alone 25 years ago when the movie came out. Not to mention - there are a host of glaring continuity errors. Joshua, for example, speaks in the same voice at NORAD that “he” does through David’s terminal, even though David himself tells us that the computer isn’t “really” talking, it’s just being modulated by his own voice synthesizing system. Since that system either doesn’t exist or would be different at NORAD, this seems hugely unlikely. And then there’s the bit about the computer going straight from Tic-Tac-Toe to “Global Thermonuclear War” when “learning” that you can’t win any games. How? No one told it to play “Global Thermonuclear War” with “number of players zero,” or to play any games other than Tic-Tac-Toe, for that matter. And while we can believe that a computer can run through all the possibilities of Tic-Tac-Toe in a matter of seconds, there’s simply no way it can do the same for “Global Thermonuclear War.” There are way too many variables. Chess alone (one of the games Joshua “skipped”) has 10¹²⁰ or so possible games - it’s a good bet that “Global Thermonuclear War” has many more. Anyone honestly think a computer can go through all of them in about a minute? More to the point, if we granted that it could, then the story itself is implausible. Joshua/WOPR’s full-time job, after all, it to do exactly that. If it had been doing its job, it would’ve reached this conclusion a long time before David Lightman hacked it.

Alright, so Hackers is technically more “real” than WarGames. Why does it seem the other way around?

That’s today’s big lesson in fiction writing, kids. It’s not enough to tell a convincing story; half of the craft is in selling it. I’m reminded of the part in Stranger in a Strange Land where Michael Valentine Smith is trying to earn a living as a magician and doing a poor job at it. Which is ironic, since as a telekinetic from Mars with spatial awareness of more than just our three dimensions, Mike actually can make things disappear for real. And yet he doesn’t really impress the crowd. I always thought this was one of Heinlein’s more convincing bits of insight. Magic is only 40% about having clever tricks: the other 60% is all in the show. Cleverness is important, but showmanship counts for more. So it is with movies too, apparently.

And dear ol’ Roger Ebert seems to pick up on this too. Here are some choice quotes from his reviews for each movie.

For Hackers:

The movie is well directed, written and acted, and while it is no doubt true that in real life no hacker could do what the characters in this movie do, it is no doubt equally true that what hackers can do would not make a very entertaining movie.

In other words, Ebert isn’t convinced, but then, he argues, he doesn’t hafta be. The classic defense: “just tell me a good story!”

Interesting, then, what he says about WarGames:

The movie, however, could easily go wrong by bogging us down in impenetrable computerese, or by ignoring the technical details altogether and giving us a “Fail Safe” retread. “WarGames” makes neither mistake. It convinces us that it knows computers, and it makes its knowledge into an amazingly entertaining thriller. (Note I do not claim the movie is accurate about computers — only convincing.)

HA! What a nice way to put it. It’s not just that WarGames fails to be accurate. It’s way beyond that, actually. So isn’t it interesting that it nevertheless manages to be convincing? How?

Well, that’s the magic, of course. If we knew the answer, then presumably someone could write a “Scriptwriting for Dummies” book that would actually work and save us from all Hollywood’s tripe. I don’t know the full answer; I don’t think anyone does. But here are some things that occur to me in this context anyway. So here goes - some tips on how to make a hacksploitation flick and get it right.

(1) Show us a hack that isn’t magic. It’s worth noting, I think, that the only hack the audience can follow in Hackers is the social engineering hack that gets our hero into the cable company early on. He tricks a hapless hourly into giving him the number on a crucial modem. From there on, of course, it’s all black magic. YAAAWWWWNNN. There’s really nothing clever to see here. The “fraud” technique used is as old as the hills. If we were pulling this same trick before the computer age, the method would hardly change. We’d still need to mysteriously know enough about the company’s organization to know exactly which dude to call, we’d still need him to be naive, only back in the good ol’ days instead of ordering the computer to play us the video we want, we’d just have to fake some instructions from upper management. Big deal. Contrast this with the big hack in WarGames. In WarGames, David Lightman programs his computer to call every number in Sunnyvale and remember which ones answer with a computer tone. Not only is it within even the most computer illiterate’s grasp to understand how this might work, it’s a clever sort of solution that you can only do with a computer. Unlike Dade in Hackers, who is playing an old prank with a new tool, David’s playing a prank that wasn’t even possible before home computers. That’s crucial. More to the point, though, WarGames is light on “black box” hacks. In Hackers, we see a lot of people typing at a lot of keyboards and things happening as a result, but the mechanisms are all mysterious. We just have to trust that these people can do the things they can do, and that wears thin after a while. It’s like showing a wizard’s duel and asking us to understand when the characters start to run out of magic. Hard to relate, since none of US can cast spells and have no frame of reference for what’s “tiring” and not here. WarGames never strays from the path on this one. Yeah, sure, we can’t exactly see David’s (BASIC? *snicker*) code, but nothing that he does requires all that much suspension of disbelief. Logging in to the school’s computer to change his grades given the week’s password? Check. Dialing every number in Sunnyvale to find Protovision? Don’t know how he did it, but the point is I believe he could. And from there it’s really out of his hands. There’s no false tension built up by asking us to watch a duel in a medium we do not and CAN NOT have any familiarity with.

(2) Real characters. I would guess that Hackers‘ biggest downfall in the computer believability department probably has little to do with actual computers at all. It’s just that these kids are all obnoxious, unlikeable, and - fair or foul - just don’t really jive with our stereotypes of what hackers are like. They come across as ravers who happen to have hobbies in computers. Everything about them is so over-the-top that it’s hard to shake the impression that the moviemakers are trying to hedge the inevitable “not realistic!” criticism from real-life hackers by hanging flashing neon signs everywhere that say “this isn’t really happening!” Contrast that with David Lightman. There’s a great scene where the FBI agent is talking about how Lightman “fits the profile” for espionage. He’s “intelligent, an underachiever, a loner, bad grades in school.” Heh. Now THAT’s a hacker! Espionage isn’t the only profile that fits. And the people making the movie know that, obviously, which is why they included that line at all. More importantly, the romantic tension between David and Jennifer works as well. David’s outside the normal jock dating rituals, not a very masculine guy. Girls aren’t a priority for him because they can’t be. So what kind of girl does such a guy get? Exactly the kind of girl that Jennifer is, obviously. Shy guys need the girl to sorta make the first move, and Jennifer is that girl for guys like David. She’s popular, probably, but seems likely to be bored with her normal options. She can party with the party crowd if she wants to, but sex isn’t really her thing, and she’d like someone with a bit more depth that she won’t have to share with everyone else. It works. So what is going on in Hackers? Despite what Ebert seems to think, cliche city, actually. It’s the standard story about a guy who gets a bit intimidated by a girl who might just be better than him at “his thing.” She’s aloof, untouchable, he bides his time, make his move by showing her up at her own game and … YAWWWNNN. Yeah, got it. This is the romantic subplot of every *sploitation movie on the market. But even more important, I think, are the background characters. In WarGames, aside from possibly David’s parents, everyone seems like a real human. They’re neither good nor bad, really, they’re just sorta pluggin along doing their thing. The debate between Mr. McKitrick and General Barringer about whether to completely automate the missile launch procedure is intriguing because we don’t really know which side is right - both men make good points. And as for Stephen Falken, he has the right idea about war, but the wrong solution to the problem. It’s a believable position for a brilliant researcher out of the 60s who’s lost his son to take, and demonstrating why it’s wrong serves a real thematic purpose in the movie. But in Hackers? Everyone is an annoying cartoon. The oil execs are conveniently hapless, the cyber criminal is conveniently a prick, the FBI guy might as well have “Moral Majority Certified Witch Hunter” stamped on his forehead, the group of hackers all act like bad high school stoner stereotypes, and so on and so on. The only bit of character innovation we really get is that there’s a Puerto Rican in this movie. Way to go out on a limb, there, guys.

Graphical Restraint. Hackers really pulls out the stops on the computer imagery stuff. If it doens’t exactly take it up to 11 like The Matrix did, it still gets a comfortable 10. There’s too much neon and techno dayglo, and it only takes the average viewer about 20seconds to get really sick of it. In fact, in my case, I can honestly say it’s the biggest barrier to my ever seeing this film again, superficial though that may sound (and be). I don’t like the style, and I can’t get away from it. So Done. WarGames looks like real life for the most part. Aside from the scenes at NORAD headquarters, there isn’t really anything in the movie that looks too technological. David’s computer equipment was dated even when the movie came out, and in any case we don’t get to see too much of it. This is a movie about the computer age, to be sure, but it mercifully doesn’t feel the need to make a fetish out of it. Unlike in Hackers, all the images we see in WarGames are things we’ve either seen before or can easily extrapolate from things we’ve seen before. There are no endless lines of purple code scrolling implausibly quickly across people’s faces.

Well, that’s my stab at it, anyway.

To be perfectly honest, I liked Hackers a wee bit better than I’m letting on here. Yeah, the romantic rivalry thing between the two main characters was cliched and irritating, and I really wanted the movie about computer culture they advertised rather than the one about rave culture I got instead. But it’s not as bad as it might have been. The central villain is interesting for being not too different from the kids we’re supposed to take as the “good guys,” which casts the whole story in a nice amoral light. Our villain isn’t doing anything these kids won’t try themselves someday - he’s only “bad” because circumstances have forced him to pick on their friend. Something to think about. Also, however unconvincing the final product turned out, it’s clear that they hired some real hackers for background research. I appreciated that the Dragon Book made an appearance, for example, and with the right cover and everything! And OK, the reference to RISC architecture was dopey, but hey, I’m pretty sure the scriptwriter doesn’t know how to fix that! Most importantly, I appreciated the general lack of sex. To the extent these characters do it at all, it’s for fun and not really an obsession of theirs. If this movie got nothing else about hacker culture right, at least it stayed true to that one. It isn’t that hackers are asexual people, but they’re certainly less obsessed with sex than the general population, and it’s nice to see that come across in a teen movie. WarGames did it better, of course, but then, WarGames did everything better.

So I’ll see Ebert’s three stars and raise him negative half a star. Hackers wasn’t good, but neither was it terrible. Tone the graphics (and wardrobe) down a bit and I might waste another two hours on it someday.

Which still gives it nothing on WarGames, of course. I’ll be watching that movie again and again till the Singularity.

I’m not usually one for paparazzi reports, but this is some very encouraging news. It seems Matt Damon was spotted at a coffee shop test-driving a Tesla Roadster, which I have blogged about before. It’s to be the first “cool” green car - a $90k sportscar (actually, a variant of the Lotus Elise) that runs entirely on battery power. At the equivalent of about 135mpg, it’s certainly efficient. The only thing keeping the rest of us from buying one is the price.

The report is that Damon is impressed and will probably buy one. I hope so. The more celebs seen in these things, the “cooler” they become, and Tesla is actually banking on the cool factor for reinvestment capital in a more affordable line.

The market will provide. The link goes to an article in the International Herald Tribune about the rising crime of stealing used cooking oil from fast food joints.

Yes, cooking oil.

Outside Seattle, cooking oil rustling has become such a problem that the owners of the Olympia Pizza and Pasta Restaurant in Arlington, Washington, are considering using a surveillance camera to keep watch on its 50-gallon grease barrel. Nick Damianidis, an owner, said the barrel had been hit seven or eight times since last summer by siphoners who strike in the night.

“Fryer grease has become gold,” Damianidis said. “And just over a year ago, I had to pay someone to take it away.”

Why? Well, it seems that fryer grease is actually a commodity - as in traded on the market - and that its value has increased from 7.6 cents a pound in 2000 to 33 cents today - which is to say it’s up over 400%. Who would’ve thought that buying fryer oil futures was even an option, let alone a good idea? But there you have it. An ambitious fryer oil thief - one who hits several restaurants in a night - can rake in $6000 on a typical haul on the black market.

Why the sharp increase in value? That’s the interesting part. It turns out the fryer oil is a legitimate biofuel. You mix it with alcohol in a process that any old boob can duplicate in his home, and cars will run on it. So what we have here is yet another example of the efficiency of markets. Gas prices rise sharply, and suddenly there’s a market in alternative fuels, even in the most unlikely of places. It’s like that old story about the wind and the sun. Government bureaucrats can spin as many schemes as they want to try to get people to go green - but what actually works is incentive, not coercion. If gas prices rise, people will go green on their own - no socialism required. Oh, and this is probably good news on the obesity front as well - since the added security measures to protect waste frying oil is bound to drive up the price of fried chicken bits.

Well, well. You know all those arguments that legalizing prostitution would lead to some kind of explosion in the rates? Not true in New Zealand, where “a comparison between the number of sex workers in Christchurch in 1999, before decriminalisation, and 2006 - after the act was passed - showed the total had stayed about the same.”

More goodies:

Before the act, the illicit status of the industry meant workers were open to coercion and exploitation by managers, pimps and clients. Research indicated there had been “some improvement” in employment conditions “but this is by no means universal”.

In other words, laws banning prostitution actually work against prostitutes by making them complicit in the “crime.” They are reluctant to go to the police with abuses for fear of arrest. Who knew?

Not to mention - sex work isn’t all (or even mostly, or even at all in many cases) about exploitation.

Around 93 per cent of sex workers cited money as the reason for getting into and staying in the sex industry.

“The most significant barriers to exiting are loss of income, reluctance to lose the flexible working hours available in the sex industry and the camaraderie and sense of belonging that some sex workers describe.”

So sex work is a good part-time job, actually. That’s backed up by this UChicago working paper, which reports that prostitutes in Chicago work an average of 13 hours a week for $26-31/hour. Better money than most of these women can earn anywhere, with flexible hours. One of the women in the study, in fact (the one with a pimp) was earning more like $50/hour for about 12 hours a week.

Is there any reason at all why this should be illegal? Oh, right, the danger of violence. Certainly that comes with the territory, but making anything against the law guarantees it will be more dangerous. Again from the New Zealand article:

More than 60 per cent felt they were more able to refuse to provide commercial sexual services to a particular client since the enactment of the law.

Other findings included that the majority of sex workers felt the act could do little about violence that occurred, although a significant majority felt there had been an improvement since the passing of the act.

So violence remains a problem, but sex workers now feel more in control of the situation. A step in the right direction rather than a panacea, in other words - exactly what one might expect.

I realize that New Zealand is a demographic outlier in the world and that its experiences with policy implementation can’t be taken as definitive for that reason. But these results are just intuitive; there is no reason to believe the situation after decriminalization or (better still) legalization in the US would be any different. Let’s legalize it already.

If This is to be believed, the Tesla Motor Company will be introducing a $98,000 all-electric sportscar soon - possibly next year. Some circles claim that it will run at an average cost of $0.01/mile, do 0-60 in 4 seconds, and require only 3.5 hours to charge (though not from a standard wall socket - it will need special “fueling stations” of its own).

It’s the 1cent/mile fuel cost that strikes me, of course. That’s really spiffy!

So where’s the catch?

Well, the price tag, obviously. By comparison, I drive a 1999 Maxima that I bought used for around $15k. Taking the current mileage and comparing it to what it was when I bought it and dividing by the number of years, it seems I drive about 6,000 miles a year. OK - ‘99 Maximas get about 21mpg (that feels about right to me). So that implies I use 286 gallons of gas a year. Since I have to fill up with premium, I’m currently paying $4.20/gallon. All other things being equal, I would drop that to an even $4 to factor in this year’s cheaper months - but since I expect gas prices to continue rising a bit, let’s just do the math at $4.20. Assuming a solid rate of $4.20/gallon, I could expect to pay just over $1200 in gas a year. Since my total expenditures on gas for the Tesla Roadster would be only $60, I would save $1040 or so a year in gas by buying one.

Yeah, but it costs so much more than my current auto. Even if we buy these numbers (and let’s face it, we don’t: I find it hugely implausible that there are cars in existence that run on $0.01/mile, not to mention that new technology like this that expects to sell only 10,000 units or so exclusively in California in its initial offering is bound to require frequent and expensive repairs), it would take me over 80 years of driving to break even (a $15k car vs. a $98k car).

The irony of all this, of course, is that people who are in a position to shell out $90+k for a car probably aren’t sweatin’ the $1000/yr gas bill much. So I don’t really see the point in talking about fuel efficiency for an electric-powered luxury car.

Fortunately, Tesla gets this. From a post by the CEO on the company’s blog:

Almost any new technology initially has high unit cost before it can be optimized and this is no less true for electric cars. The strategy of Tesla is to enter at the high end of the market, where customers are prepared to pay a premium, and then drive down market as fast as possible to higher unit volume and lower prices with each successive model.

Without giving away too much, I can say that the second model will be a sporty four door family car at roughly half the $89k price point of the Tesla Roadster and the third model will be even more affordable. In keeping with a fast growing technology company, all free cash flow is plowed back into R&D to drive down the costs and bring the follow on products to market as fast as possible. When someone buys the Tesla Roadster sports car, they are actually helping pay for development of the low cost family car.

Needless to say, this is encouraging, a really smart strategy.

Ah, but do the cars work? Popular Mechanics seems satisfied!

I will be needing a new car in exactly two years. If they have family cars on the market by then in the $50k range, I’m in.