Thatcher's rants and musings 2005

# permalink

1 Dec 2005

The Power of Cheese

I have a simple recipe for killing mice:

1. Start with lots of mice. Our household is currently weathering a flurry of mouse activity, so we've got that covered.

2. Buy some ordinary mousetraps. The kind made of a piece of wood and some wire, invented like 100 years ago.

3. Bait with a little cheddar. I started with peanut butter, but the mice around here seem to greatly prefer cheese.

4. Place them somewhere the mice frequent, but the dog can't get to.

5. Wait a couple hours. Results:

Speaking of the dog, you might think Pokey would clear out the mice on her own. After all, she's OBSESSED with slightly larger rodents like squirrels, rats and bunnies. She sits at home bored most of the day. She'll eat practically anything that's gross, and dead mice certainly qualify. In the house though, it seems she is afraid of mice!

# permalink

6 Nov 2005

Textweb does Atom

Textweb can now generate Atom XML feeds of blog-like content. For example, here's this blog rendered as Atom.

# permalink

29 Oct 2005

Hazel has reached an important developmental milestone: TV obsesssion. Up til now she's been curiously resistant to heavy TV watching -- she would pay rapt attention to five or ten minutes of Teletubbies, but then wander off to do something else. This has been an extra burden on our babysitting duties, but on the other hand she's always been an exceptionally good sleeper.

I was beginning to worry about the TV thing. But the tide shifted this week, culminating in today's bonanza -- she somehow managed to squeeze in at least four fully attentive viewings of the Baby Einstein "Shapes" video, and one or two Dora episodes, in addition to tolerating the miscellaneous news and E! True Hollywood Story her parents are addicted to.

Go Revs

The dreaded sports blog, I'll try to keep it short. The New England Revolution made a rousing comeback to win their first-round playoff series against NY/NJ. The Revs dropped the first game 1-0, so they had to win today by two goals to take the series. They went down 1-0 again, early in the second half, but Jose Cancela came off the bench and rescued the team with a goal and an assist. The Revs ended up winning the game 3-1, and the series 3-2.

I wonder if Cancela will start the next game. He ought to; since Daniel Hernandez displaced him, the team hasn't been as good, IMHO. Hernandez is a good player, but I think his strengths are too redundant with Shalrie Joseph's.

# permalink

16 Sep 2005

LuaJIT

LuaJIT is very interesting -- it's a just-in-time compiler for Lua, courtesy Mike Pall. It's very small, like Lua. Based on feedback from the Lua mailing list, it's very robust and fast as well. If you're planning to embed a scripting language in a PC game, this tips the balance heavily in favor of Lua, IMO. It's only for x86 so far though. But Mike claims it's not hard to add new backends.

# permalink

25 Aug 2005

The Psychology Of Computer Programming

I just read this book, by Gerald Weinberg, originally published in 1971 or so. It's considered a classic, but I'm not sure I recommend it. Peopleware by DeMarco and Lister is denser, more entertaining, and more up-to-date, so read that instead. I found this one very slow going at the beginning, and kind of windy throughout.

I learned a few interesting things though.

For example, he talks a lot about the dynamics of batch processing, from back in the day when programmers would submit a punchcard deck to a machine operator, and come back hours or days later for a printout.

Apparently, intensive code reviews were typical for corporate coding teams in the batch days -- it was just something you logically did if you wanted to be productive, since the smallest syntax error would sabotage a whole batch run. So coders learned to read programs, and code got reviewed.

It seems that interactive timesharing and the PC era more or less killed off that tradition of code reviews, and so the younger generation has had to rediscover its benefits.

There are some miscellaneous factoids and principles from psychology, some of which are pop psychology by now:

The Hawthorne Effect.
The theory of "cognitive dissonance" -- people tend to irrationally reject evidence that threatens self esteem or contradicts deeply held beliefs.
Theory of learning -- different people have different learning styles, etc.
"Driving force" vs. productivity: increased driving force increases productivity to some maximum level, beyond which performance quickly dwindles to zero (especially for complex tasks). He makes the interesting distinction between "inner driving force" and "outer driving force" -- and that many programmers are overmotivated to begin with, so that applying much external pressure often causes projects to fall apart.
"Egoless programming" -- I think Weinberg coined this phrase, and makes a good case for it here.

There's some reasonable stuff about scheduling and tradeoffs and communicating with management. There are a few corny jokes and dated stories about miniskirts I found amusing.

# permalink

13 Aug 2005

Today's Ups And Downs

Up: we arrived in the Vineyard yesterday for a week's vacation.
Down: New England is experiencing a record heat wave, with the usual zillion percent humidity. We have no air conditioning in our creaky old house here, so everybody is hot, sticky and cranky.
Down: We left the rental car on the mainland and took the ferry on foot, since we leave a car over here. But our sweet 16 year old Volvo doesn't start, and it's Saturday so we can't get it fixed.
Up: Things could be worse; we can walk to the beach (we did) and to downtown Oak Bluffs for lunch (we did).
Down: I volunteered to bike to the farmer's market and get some good fruits & vegetables. I got about two miles away from home before I discovered that both bike tires were rapidly losing air, so I had to walk back in the midday sun on a narrow road, empty handed.
Down: I remembered to bring a laptop, but managed to forget to bring the AC power supply. I spent a couple hours today, first on the phone to the two local computer stores (neither of which carry Thinkpad power supplies). I called my employer's helpdesk, who didn't think my situation was worth any extraordinary measures (well, they're right about that, it's not like I'm going to do any work this week if I can help it). Then, on waning battery power, I tried finding some online vendors of Thinkpad power supplies who might answer their phone and overnight me one -- but apparently in this day of e-commerce, nobody answers their phone on Saturday anymore. Except this one guy, Bill Morrow who runs http://thinkpads.com -- he picked the phone right up and offered to help me out if I could get FedEx to make a late Saturday afternoon pickup at his place in Florida. No dice.
Up: The thinkpads.com guy did point out that the Thinkpads almost all take the same simple 16 volt supply so maybe I could scrounge something. I figured Radio Shack might be the ticket. With both car and bicycle out of commision, my only viable option to get there was my electric motorbike, whose batteries have been unused and untested for the past year. I took the chance, the Voloci worked great, I got a bunch of produce at the farm stand, and I stopped at Radio Shack while I was out. I figured I might get to cannibalize the power supply to some toy or something and solder on the right connector for my laptop. Instead I found the "iGo wallpower" for around 70 bucks, which is a universal laptop power supply. Moral: I love Radio Shack, I hate all you loser online laptop stores (except for thinkpads.com!).
Down: At the fancy fruit stand, I spent around $3 on a single nectarine for Julie, but purportedly a very special one, by the brand name "Honey Drop". It turned out to be dry and mealy!
Up: As we were grilling burgers, a neighbor dropped by and gave us a big bowl of freshly harvested clams! So we had burgers and steamers for dinner.

Shoddy Embedded Software

The sloppy reliability of PC software is invading the embedded world. Things in my life that spontaneously stop working, but can be fixed by a power-cycle (i.e. physically pulling the plug), include:

Our home wireless router thingy, a Netgear WGT624 I think. Don't buy this one.
Our DSL modem, though it hasn't fritzed out for months; somehow it has become more reliable over time.
Our Pioneer/Time Warner cable box.
Apple Studio Display (a really old one, but it's been flaky this way since the first week I owned it)
Julie's Apple-brand bluetooth keyboard and mouse peripherals (brand new).

Embedded developers, here is my advice:

First, as a simple failsafe measure, you need to learn about watchdog timers, e.g. see this article: http://www.embedded.com/2000/0011/0011feat4.htm
Second, stop writing brittle software in the first place.

# permalink

30 July 2005

Jury Duty

I just finished two weeks of jury duty, on an actual trial. I'd never been selected for a trial before. I found it partly interesting, but mostly excruciatingly boring. The thing is, the case was just not dramatic -- a civil matter involving an accidental warehouse fire on the Brooklyn waterfront, that consumed about a million bucks worth of cocoa beans imported from Cote d'Ivoire. The insurance company for the cocoa trader accused the warehouse managers and the warehouse landlord of negligence. The fact that the cargo was destined for chocolate bars was the most appealing part of the whole affair, but that wore off after a day or two of mind-numbing testimony.

The legal process was interesting to see though. We got to see the rules of evidence in action, and a parade of characters giving differently-skewed versions of the same events. In the end, I think we gave the right verdict: the negligence of the defendants substantially contributed to the fire. Welders repairing the warehouse at the time probably caused the fire by a carelessly discarded cigarette. The landlord's supervision and oversight of the work was shoddy at best, and the warehouse management wasn't paying much attention either.

Some notes:

We were a very young jury; at 36 I was the oldest by a decade or so.
Everybody called me Walter throughout the whole thing, since that's what's on my jury card and it was too much trouble to explain that I go by my middle name.
They put you in an impossible situation: you listen to the case for hours, and then for a break, they put you in a room with the other jurors, and tell you "do not discuss the case". For two weeks I had the worst jones for talking about the case, and no way to release it. But as soon as we delivered the verdict and could talk about it, I just didn't care very much anymore.
During the trial, especially at the beginning, I was worried about some aspects of deciding the case, like that it would be hard to remember relevant testimony, separate my personal feelings about the parties and their lawyers from the merits of the case, filter out spurious arguments, and ignore interesting tidbits that had been testified, objected to and then "stricken from the record". But, in the end I feel that none of those things were real problems. It takes so long to get through everything, that you have more than enough time to chew things over and decide what you believe and what matters. Having six independent jurors hash things out at the end seemed to work pretty well.
A strange factoid: "cigarette filters will survive a fire that consumes an entire couch". That was roughly the testimony of the defense's expert witness, though it seems wildly improbable. True or false? (Fortunately this issue was not actually relevant to the case.) I've been searching the web for confirmation or contradiction, and haven't found anything good yet. I have learned that most cig filters are made of cellulose acetate, which is essentially rayon. I'm sure cellulose acetate will melt and presumably burn if you get it hot enough, but I imagine cigarette filters are treated with flame retardants or something. I need more input on this. If I can't find a good answer, I might need to do some experiments!

The main defense attorney's opening and closing arguments were classic examples of asserting the opposite of the truth. He said the plaintiff's case was smoke and mirrors and we should be skeptical; unfortunately that was a bad strategy for him: when we applied the same skepticism to both sides, the defense was far shakier.

The defense attorney may have believed what he was saying, and he certainly wanted us to believe it, so he wasn't being intentionally ironic. But I notice the pattern more and more -- when somebody is trying to convince you of something, the opposite of what they're saying is often a pretty good hypothesis. Is there a word for this pattern? Maybe "doublespeak". Examples:

when you see a sign in a restaurant window that says "Best Burger in town!" you can be pretty sure that their burger sucks.
"We think you will find jury duty an enjoyable and fascinating experience!"
"My dog is friendly."
"In order to serve you better, we require the following information..."

# permalink

13 July 2005

The Terrorists Are Winning

In last week's New Yorker, Jane Mayer writes about our prison camp in Guantanamo Bay. Link: http://www.newyorker.com/magazine/2005/07/11/the-experiment-3

Reportedly, the US Government has been torturing and abusing prisoners there for the past several years, with explicit approval from the Secretary of Defense and the White House legal department. We have used techniques such as:

physical beatings
withholding of medical treatment (e.g. for broken fingers suffered during beatings, among other things)
threats of violence
long term sleep deprivation
long term isolation
numerous tactics of psychological abuse -- desecrating religious symbols, humiliation, sexual taunting, continuous loud noises (e.g. a loud tape loop of a crying baby playing 24/7)

These could be straight out of Solzhenitsyn's book The Gulag Archipelago, that chronicles the atrocities committed by the Soviets in prison camps following WWII. See this excerpt, for example: http://207.44.245.159/article9236.htm

(In another parallel to the Soviet Gulags, many of those imprisoned are innocent of any crimes against us. E.g. see this LA Times article from 2002.)

To the above forms of torture, we have added our own particular innovation: "scientific" reverse-engineering, by medical and psychiatric personnel in our military, of anti-torture training conducted by our military. In other words, our military has a program to train our own soldiers and pilots to resist interrogation in case of capture, essentially by conducting some of the above types of torture under controlled conditions, and measuring the results. The twist is, some of the scientists involved with this program have been using their data and experience in this program to help devise interrogation techniques to use against prisoners at Guantanamo.

Because we've been torturing these prisoners at Guantanamo, they can never be prosecuted through our real legal system, no matter what their crimes might have been and what evidence we have against them. As far as I know there are only two possible outcomes for such a prisoner:

death while imprisoned -- their lives taken without trial or due process
release to their home countries, where they will not lack reasons for wanting to kill us

I'm not sure what else to say about this, hyperbole does not do it justice. We, as Americans, cannot claim that we don't know that this is wrong. The various legalistic arguments used so far by our government to justify it don't stand up to reality. All I can think is, our system is cracking -- we are helping the terrorists beat us.

# permalink

23 June 2005

The Pragmatic Programmer

Lots of people love this book. I just finished reading it. Actually I skimmed it -- I didn't find much to disagree with, but on the other hand I didn't find much that surprised or excited me. Notes:

They love automatic code generation. Granted, it's a very powerful technique, but there's a big caveat they left out, if we're trying to be pragmatic. Which is, code generation can be a major problem to code reuse, especially open source. The reason is kind of stupid, but true: most coders are stuck with bad/inflexible build systems that they don't understand, and those people will have a hard time hooking up your generator.
The one piece of useful insight that surprised me was tip 69: "Gently Exceed Your User's Expectations". As opposed to greatly exceeding their expectations. I had always thought more exceeding is better. But, they point out that if you wildly exceed expectations, you probably did work that wasn't necessary, or spent effort that might have better gone elsewhere. And, you didn't properly communicate what you were going to deliver, as you were developing.

WebTweaker

A good gamedev practice is to have an easy-to-use system for tweaking live parameters in the running system. People do it with heads-up-displays, command-line consoles, GUI widgets, both local and remote. For a long time I've been meaning to code up a reusable one for tu-testbed, using a custom GUI treeview talking to the game via a socket. I realized today what I really want to do is make the game a webserver, and use a browser as the remote tweaker app. It should be possible to make a good web interface for tweaking, and it nicely eliminates the problems of client platform dependence, building a good communication protocol, and coding the client. Props to The Pragmatic Programmer for reminding me that this is handy.

# permalink

15 June 2005

FYI

What I've been up to lately:

Working at Google, so far so good. I have to admit, I miss graphics programming a little; it would be nice to not feel like a dumbass all the time.
Flying cross-country with a toddler. My advice is, don't do it. If you do, bring plenty of jellybeans and Swedish Fish. Hazel behaved amazingly well under the circumstances, but even so I ended the "vacation" far more exhausted than when I started.
Trying to unstuff my ear after said plane flight. After touching down in NY with severe sinus congestion, my ear remained underpressurized and unpoppable for nearly a week! It felt kind of like my head was wrapped in a towel, let me tell you that's a very nice feeling in an NYC heat wave.
Pondering Robert Glass's "Rules of Three in reuse", from his book Facts and Fallacies of Software Engineering.

"Fact 18: There are two 'rules of three' in reuse: (a) it is three times as difficult to build reusable components as single use components, and (b) a reusable component should be tried out in three different applications before it will be sufficiently general to accept into a reuse library."

Some other guy has already blogged about this very topic.

In my experience, the Rules of Three sound right. Glass gives some other interesting reuse facts, here are a couple highlights:

"Fact 15: Reuse-in-the-small (libraries of subroutines) began nearly 50 years ago and is a well-solved problem."

"Fact 19: Modification of reused code is particularly error-prone. If more than 20 to 25 percent of a component is to be revised, it is more efficient and effective to rewrite it from scratch."
I think I can render SWF's filled quadratic bezier shapes directly, without doing any tesselation or curve subdivision. The idea is to use stencil along with a quarter-circle alpha-tested texture (or shader function), to handle the curved parts. I haven't coded it yet, but the details are in these notes if you're interested.
Thinking random thoughts like: if the morons spreading phishing attacks could spell, they'd be much more dangerous!

# permalink

10 May 2005

Suggested License Plate Caption

California: It's Still Shiny.

# permalink

25 April 2005

Zipcar: The Big Lie

So the thing they didn't tell me before I signed up for Zipcar is that I cannot get a car in NYC on the weekend. The nearest unreserved Zipcar this Saturday is 60 miles away, in Princeton, NJ! Same story for the next EIGHT Saturdays!

Well OK then, I guess it's back to AVIS.

# permalink

8 April 2005

NYC Rants

I guess there's no spring this year. One day it's freezing rain, then we have one perfect day, and the next day its that old familiar sticky/muggy feeling.

I'm developing sidewalk rage. Why do so many people walk so slowly???

# permalink

31 Mar 2005

Oddworld

Charles reports that Oddworld has ceased game production operations (See entry for 3/30/05). I haven't announced it here before, but I left Oddworld (and the game industry) last month for unrelated reasons, so this doesn't directly affect me. Still, this is a big disappointment. The Stranger engine is good, and would have been a great base for more games. It's obviously going to be a big disruption for current Oddworld employees. I don't really know anything more than what's in Charles' note and the gamespy article, so it's hard to draw any firm conclusions about why this happened.

Oddworld was a great job for me; I learned a ton and it was a really good team. I think it was one of the better-run game companies around.

I'm becoming ever more cynical about commercial game development of any scale though. I think innovation is games is more likely to come from amateurs than from anyone else.

# permalink

17 Mar 2005

IGJ3

I participated in IGJ3 a couple weeks ago (official site at http://indiegamejam.com but stuff from this year hasn't been added yet). This year the theme was, "People Interacting" (or something like that), and we had access to The Sims (version 1) character models, textures and animations, which was amazing. We also had some great voice acting & extra sound support this year. (The sound in previous years has been great too, but this year we had more personal and it got more elaborate.) The theme proved to be a challenge. The crew and space were larger this year, but it seemed like there were fewer playable games at the end of it. Personally I got badly bogged down in tweaking animation/pathfinding, and didn't have a nominally playable demo until mid-day on Sunday, which means my game isn't a game, more of a very rough sketch.

A bunch of people finished and did cool stuff. Ranjit stole the show with a "Waiting For Godot" game, with inspired (non-)gameplay. cbloom did a hilarious "The Office"/"Office Space" kind of Software Manager game. Chris Butcher did a slick/creepy High School social status sim. Etc.

I have lots more to say about IGJ3. I'll probably post again when the official site gets updated.

# permalink

21 Feb 2005

Sunja's Radish Kimchee

This stuff is insanely delicious! Made in Vermont. I get it at Whole Foods. Although, as Julie will attest, the smell it leaves behind is a cross between sewer leak and dead rodent.

# permalink

17 Feb 2005

Nighttime is the Right Time for The Gates

A couple of good developments:

This week's downpours have washed out the pleats. Where the fabric clamps into the crossbar they're still there obviously, but they disappear within a few inches. If you don't look for them, the pleats can be forgotten.
I took a walk through the park at night. Human color perception is drastically reduced in low light conditions, so at nighttime the color is fine. It makes me wish I had never seen The Gates in daylight.

Last night for a few minutes, in the streetlights and moonlight, a few stars out, no helicopters, no pleats, a good breeze going and nobody around, the gates finally looked beautiful.

# permalink

14 Feb 2005

Gates Thoughts

The Gates are disappointing. The scale is more redundant than impressive or transcendent. The color is ugly, except in certain wind/light conditions. People say it looks like a construction site, and people are right. The regular, mechanical looking pleats are ugly. When the fabric is billowing in the wind, the color improves and the pleats disappear, so that's good. Unfortunately the fabric seems to be too heavy; mostly it just hangs there. The other problem with the billowing is that it becomes apparent how short the drapes are, relative to the gates themselves, and it looks kind of pathetic.

If I were Christo or Jeanne-Claude, I would be deeply depressed right about now.

The concept is exciting, the logistics are flawless, and I really want to be moved, but it's just not happening on the level I was hoping for. So far I've gotten the most enjoyment under these conditions:

a sequence of exposed gates with moderate wind; the fluttering drapes catch the light and look kind of neat.
looking at something above and behind a nearby sequence of gates (i.e. focus on some buildings or trees in the background; don't foveate on the orange color or the pleats). The color drops out a bit and the drapes start to look more serene.

The Gates might look good at night, with less color perception, maybe I'll try that.

# permalink

12 Feb 2005

Unfurling The Gates

They're dropping the drapes right now... I took a few photos while walking Pokey. I have a couple initial thoughts:

Pleats?
The din from helicopters is a little bit distracting.

# permalink

8 Feb 2005

The Gates

Today (hm, or yesterday I guess) workers started putting up Christo & Jeanne-Claude's "The Gates" project in Central Park. Like many other Christo projects past, the concept seems kind of goofy. The dark gray bases and cheesy orange markers have been in the park for weeks now, with no apparent progress. But now that the gates are going up, walking around and seeing the massive scale of it, I think that when they unfurl the fabric, it will be AMAZING!!!

# permalink

1 Feb 2005

Game Developer Union

A Game Developer's Union is a concept that comes up now and then. It'll never fly as long as it's called a "union"; most developers are too young, libertarian and macho for that. But call it a "guild" instead, make up some members-only shwag like logo-stamped pewter mead flagons and chainmail T-shirts, and it would have a chance...

# permalink

31 Jan 2005

Tuffy Family

Tuffy updates on my fonts page.

# permalink

31 Jan 2005

More Transit Links

"Automated Highway System" seems to be the official buzzword in the transit field, for schemes along the lines of driverless taxis. Unfortunately Congress killed the promising US program in 1998, but academia, California and/or France may be filling the gap.

Michel Parent on Automated Public Vehicles INRIA seems to be onto something, as usual.

Steven Shladover, "What If Cars Could Drive Themselves?" This is about automation of highway driving, a step in the right direction.

# permalink

25 Jan 2005

The Work Of Director Spike Jonze

A friend gave us the The Work Of Director Spike Jonze DVD, which we have been enjoying. It's mostly music videos, Weezer and the Beastie Boys and Fatboy Slim and such. Most of these videos are new to me -- at the time they were being produced I didn't have cable, nor much interest in MTV.

There is incredible stuff in here. I love Jonze's intuition in interpreting the meaning of the songs. The flaming man running down Sunset Blvd is a perfect emotional translation of Wax's "California". The Torrance Community Dance Troupe performing their hearts out to cheesy Fatboy Slim pop. The dog-faced boy out on a Saturday night, with boombox playing French house music. Weezer performing in Arnold's Drive In.

In my dreams, I am the kind of game designer that Spike Jonze is as a movie director. Unfortunately, nobody (least of all, me) is that kind of game designer yet, because of the damn technology, and the way the market and the industry operate. We don't have the game-development equivalent of the camcorder yet.

Anyway, the thing I was going to say, before I got sidetracked into talking about the Gay Ministry, is that based on one of the videos on the Jonze DVD, Hazel looks and sounds a lot like a miniature Bjork.

# permalink

25 Jan 2005

More On Cars

Last time I spewed uninformed speculation about car alternatives. Later I did some research. Here's a great website in case you want actual expert opinions on advanced transportation: Innovative Transportation Technologies @ U of Washington

Here's a good overview paper by that site's maintainer, J.B. Schneider, Emerging Auto-Competitive Transit Technologies: A Brief Review

Basically he says the three leading contenders are:

PRT or "Personal Rapid Transit". Essentially, automated personal cars running on special purpose elevated tracks.
GRT or "Group Rapid Transit". Like PRT, but with bigger cars, more like vans. Schneider notes: "A GRT system has been successfully operated at West Virginia University in Morgantown for more than 25 years." I actually rode in that thing in 1988 while on tour with the A.G's and NOFX, and it was cool.
DT or "Dualmode Transportation". This uses regular cars on regular roads, but with modifications so that they can run on special high-speed traintrack-like highways for longer distances or higher congestion areas.

They all sound pretty neat. Driverless taxis have similarities to all of these. I think driverless taxis will beat all of them to wide scale adoption, because they avoid the high cost & political coordination of building a whole infrastructure of special new tracks.

# permalink

18 Jan 2005

Cars

Terrapass For a few bucks a year, you can clean up after your car! This beats the crap out of buying a hybrid, in terms of effectiveness. (Disclaimer: this is another of my brother Karl's ventures.)

Cars, even perfectly clean ones, still suck, for lots of reasons. Sprawl, noise, dirt, traffic, collisions, breakdowns, road rage, alienation, dog incompatibility (by that I mean, cars kill unleashed dogs), etc.

If we could just get rid of cars, the world would be so much nicer. The only problem is, cars are fantastic transportation. There is really no comparable replacement. (A car is also a good substrate for a bumper-sticker, a handy place to stash dirty clothes or a toolbox, and a real nice vocal booth for diva practice, but presumably we could find good alternatives for those uses.)

Anyway, we seem to be stuck with cars. If petroleum disappeared tomorrow, there would of course be a bad economic crash, coinciding with a mad scramble to run cars on something other than gasoline. In a few years we'd all be driving battery/plutonium/rubber-band/whatever cars.

Manhattan is an interesting example of the impossibility of replacing cars wholesale. It's dense, fairly small and flat, there's good public transit everywhere -- in short, it's the perfect test case for a car-free city. Many Manhattanites do in fact not own cars. But, nevertheless:

The streets are choked with cars. And we'd starve to death quickly without all the trucks.
If you can spare the cash, a taxi is still faster than most public transit most of the time.
I hate to say it, but if you live outside the city and come here to work or for a short visit, it may make sense to drive your car in.
If you ever want to leave the city, you really need a car out there in the rest of America. Rented, borrowed, whatever, you need one.

So, is the world forever doomed to sprawl and traffic? I hope not. Some of my random ideas:

Driverless taxis. One of the big downsides of mass car ownership is parking. Parking lots suck; they encourage sprawl, consume valuable land. Street parking takes up tons of precious urban space. Driverless taxis wouldn't need to park very often. When they did, they could drive to an out-of-the-way taxi depot.

Another downside of using a car in the city is having to drive in traffic. It's just stressful. On the other hand, riding as a passenger in a car in a city is pretty nice. Same thing on long boring highway drives; I would much rather read or watch a movie or take a nap than drive.

Driver-ful taxis are fine in dense cities, where business is steady, and taxis are pretty easy and quick to hail, and trips are usually short enough to not cost an arm and a leg. In rural areas, or the suburbs, or more spread-out cities, taxi service is so spotty it's a real pain to use. Driverless taxis could get rid of a lot of the overhead and headaches of running a small taxi service in such places, and make it something ordinary people could actually depend on for regular use.

The obstacles to driverless taxis are mostly things that can be solved with computers & networks, so why not?
Safe, fast, little scooters. A la Segway, but much faster. If you could have something that safely goes about 35 mph and is light & small enough for an ordinary person to carry, I think it would make a difference. In my opinion, current scooters are not enough better than bicycles to make a big dent in the car problem.
Transporter booths, personal jetpacks, seven league boots? I thought I had more than two non-scifi ideas for how to kill cars, but I don't...

# permalink

3 Jan 2005

Sunday Night At The Movies

SNATM rides again. Actually I've been back on the Sunday movie wagon for over a year now. I've revived another tradition, the movie reviews page.

# permalink

3 Jan 2005

Tsunami

We gave some money to Doctors Without Borders. Julie trusts them and thinks they're relatively efficient. If you're wondering who to send money to, I think they are a good option.