Elementary, my dear human

On my commute recently I listened to a recording of the talk given last month by Vernor Vinge to The Long Now Foundation on the subject of alternatives to “the Singularity.”

Vernor Vinge is an acclaimed science fiction author and futurist. The Long Now Foundation is an organization of technologists, artists, and others dedicated to pondering the challenges facing society on very long time scales, on the order of thousands of years. And “the Singularity” is a concept invented decades ago by Vinge that says, in effect: technological progress is advancing almost unavoidably to a point (called the Singularity) where technology itself will exceed the intelligence and abilities of humans. After the Singularity, continued technological advancement is in the hands of technology that’s literally superhuman. It proceeds at a superhuman pace according to superhuman motives. Just as our laws of physics break down at the event horizon of a black hole, it is in principle impossible for us to make predictions about the future beyond the Singularity, when things will be as incomprehensible to us humans as, in Vinge’s words, “opera is to a flatworm.”

Although Vinge believes that the Singularity is the likeliest non-catastrophic outcome for the future of humanity (and there are many who agree and many who don’t), his talk to The Long Now Foundation addressed alternative, non-Singularity possibilities. What might prevent the Singularity from occurring? War and various catastrophes on a global scale are obvious ones. But there are two interesting non-Singularity possibilities that Vinge did not discuss.

The less interesting and less likely of the two possibilities is that there is some fundamental limit on the complexity of information processing systems, and human brains are already at or near that limit. If these two suppositions are true, then it is not possible for technology to exceed human reasoning or inventing power by a significant amount — though it would still be possible to employ vaster, harder-working armies of reasoning and inventing machines than it would be to recruit similar numbers of people. (Interestingly, Vinge posits just such a fundamental limitation in his science fiction masterpiece, A Fire Upon The Deep — a rousing and thought-provoking adventure, and the only sci-fi story I’ve ever come across that feels truly galactic in scope.)

Here’s the non-Singularity possibility I like better: though machine intelligence may exceed that of humans, human intelligence can keep up, like Dr. Watson arriving at a conclusion or two of his own while following Sherlock Holmes around, or like me surrounding myself with friends whose superior intellect and wit eventually rubbed off on me, at least a little.

Consider that a hundred years ago, it took geniuses at the pinnacle of human intelligence to devise the counterintuitive physical theories of relativity and quantum mechanics that, today, are grasped (in their rudiments) by children in middle school. Consider that the same race of beings that once gazed up at the heavens and made up fairy tales about the constellations has now charted and explained very much of the visible universe, almost all the way back to the beginning of time — and it took only a few dozen centuries.

Perhaps there are realms of thought and invention that require posthuman brainpower to discover. But I’m optimistic that where our future technology leads, we can follow.

World widescreen web

Thinking of upgrading your conventional picture-tube TV to a fancy new flat-panel widescreen? But you’re on a budget and don’t want to go overboard? Confused about what size TV to buy? You’ve come to the right place.

The main criterion for choosing a screen size is one that I have not seen described in other TV buying guides: viewing area. The viewing area of a 32″ conventional TV is 492 square inches, whereas the viewing area of a 32″ widescreen TV is a mere 438 square inches! If you’re upgrading from a 32″ conventional TV you’ll want at least a 34″ widescreen to get the same viewing area.

Here’s how I arrived at those figures.

The advertised size of a TV display is the length of the diagonal. If from the diagonal we can determine the height of the display, h, and the width, w, then the viewing area is h×w. Thanks to Pythagoras we know that h2+w2 = 322. But this isn’t enough information to determine the viewing area: we also need the fact that the aspect ratio of most conventional TV displays is 4:3, which means the width of the display is four-thirds the height.

Substituting 4h/3 for w and then simplifying gives us:

h2+(4h/3)2 = 322
h2+16h2/9 = 322
25h2/9 = 322
h = √(9×322/25)
h = 3×32/5 = 19.2

Plugging that into the formula for viewing area (h×w) and recalling that w = 4h/3,

h×4h/3 = 19.2×4×19.2/3 = 491.52 square inches

Knowing that the aspect ratio of widescreen displays is 16:9 and using similar arithmetic gives a result of 438 square inches for a 32″ diagonal.

In fact, the math shows that for a given diagonal, the viewing area of a 16:9 display will always be about 11% less than the viewing area of a 4:3 display.

But wait! It’s not as simple as finding the widescreen TV that has at least the same viewing area as your conventional TV. You should also take into account the kinds of programming you watch.

Do you watch a lot of wide-format movies on your 4:3 TV? If so, you’ve certainly noticed the “letterboxing” needed to fit the wide aspect ratio of the film into the narrow one of the display. You’re not using the entire viewing area; some of it is wasted, as much as 32% of it for very wide format formats such as “CinemaScope.” With a 16:9 TV the need for letterboxing wide-format movies is decreased or eliminated.

Similarly, if you watch a lot of conventional TV programming (sitcoms, newscasts, etc.) on a widescreen TV, you’ll get “reverse letterboxing,” also called pillar boxing, where the black bars appear not on the top and bottom but on the left and right of the image to make the taller aspect ratio fit into a shorter one. Here again you’re wasting some of your viewing area.

So think about the kinds of programming you watch and consult this handy table that shows the true image size (in square inches) for various combinations of TV diagonal size, TV aspect ratio, and programming aspect ratio. Choose a TV that gives you the best image size you can afford for the types of programming you typically watch.

Program aspect ratio
very common
some movies
20″ 192 154 144 138 109
27″ 350 280 262 252 199
32″ 492 393 369 354 279
36″ 622 498 467 448 353
42″ 847 677 635 610 480
46″ 1016 813 762 732 576
50″ 1200 960 900 865 681
20″ 128 160 171 164 129
27″ 234 292 312 299 236
32″ 328 410 438 420 331
36″ 415 519 554 532 419
42″ 565 707 754 724 570
46″ 678 848 904 869 684
50″ 801 1001 1068 1027 808

Déjà recherché

It’s weird when you go to enter a search term in Google and your browser autocompletes it because you’ve done that search before, but have no memory of doing so. This happened to me yesterday when I typed “shoel” into the Google search box and my browser provided “shoelace knots,” which is just what I was going to search for. Must have done that search once before, though I don’t remember when or why.

(I was looking for a shoelace knot that would keep my laces out of my bicycle’s chain ring. Yes, I’m finally biking again; cutting out pasta alone isn’t doing the trick.)


Recently I discovered that Sony is planning a new in-dash car stereo with a USB port for reading and playing MP3 files from a thumb drive. I had an earlier such model but it got stolen. I replaced it with a JVC model that I don’t much like. (Among other things, its clock won’t display when the unit is switched off. And when it does show the time, occasionally the time display won’t update itself for many minutes in a row!) For someone with a long daily commute, a good car stereo is indispensable. So I’m interested in Sony’s latest and I have been checking Amazon periodically to see when it’ll be available to order.

While checking just now I discovered that Harry Potter and the Deathly Hallows — the final book in the series — is now available to pre-order. Its just-announced release date is 21 July. Yay! Into the shopping cart. I’m proud to be among the first to pre-order it. Thanks to an automated recommendation I also discovered that Universal Hall Pass — Melissa Kaplan’s post-Splashdown band — has a new EP out. On it she collaborates with former Splashdown bandmate Kasson Crooker — yay! Into the shopping cart.

Today’s run of luck continues…

Clever climb clock

The rate-of-climb indicator, also called the vertical-speed indicator or VSI, is one of six instruments in the standard instrument cluster familiar to airplane pilots. It reports the rate at which the airplane’s altitude is changing, in hundreds of feet per minute. (Interesting fact: in an unpressurized airplane cabin, a comfortable rate of descent is five hundred feet per minute. Much more than that and passengers will begin to feel ear pain from the pressure changes. So a pilot flying 5,000 feet above the elevation of his or her destination should begin descending while still ten minutes away from the airfield.)

The VSI does not need a power source. It doesn’t need to be wound or filled or set. There are no tubes, wires, gears, levers, or lines of any kind leading to or from it. It just sits there, bolted into the instrument panel, requiring nothing but some clear space around it for air to flow freely. How can it possibly work?

The altimeter, another standard instrument, is much easier to understand. It registers the airplane’s height above sea level by measuring the ambient air pressure. (Interesting fact: the ambient air pressure runs to the altimeter through a tube originating at a tiny hole in the skin of the plane called the “static port.” An airplane in flight affects the pressure of the air all around it; the static port is strategically placed where the effect on the surrounding air pressure is neutral.) Air pressure decreases at a pretty constant rate the higher you go, so if you know the pressure at sea level (a setting that changes from place to place and from hour to hour — pilots periodically get the setting from a radio broadcast and adjust a knob on the altimeter), and you know the air pressure, then you know your altitude. The altimeter is nothing but a funny-looking barometer calibrated in feet above sea level.

Even the airspeed indicator, which is a bit cleverer, is easy to understand. It uses two sources of air pressure: the “static” air pressure (from the aforementioned static port), and the “ram” air pressure, which is the pressure of the oncoming air as measured by a tube (the “pitot” tube, rhymes with Frito) pointing forward. Via some simple plumbing, the static pressure is subtracted from the ram pressure and the result is shown on the airspeed dial, calibrated in knots or in miles per hour.

The design of the VSI is a whole ’nother level of cleverness. For a long time I tried to puzzle it out myself before I finally relented and looked up the answer. Can you figure it out?

Continue reading “Clever climb clock”

Today is Thursday, 2007

Yesterday I finally got around to resetting the clock on my answering machine at home after the big power failure we had recently, when I noticed this oddity for the first time: it prompts for day of the week, time of day, and year. No month, no date. As I write this, according to my answering machine, it’s Thursday, 12:14pm, 2007.

This year it will be Thursday at 12:14pm fifty-two times. Not having the month or date is a real obstacle to knowing which one is meant. And when I play back my answering-machine messages, the machine says, “Thursday. Twelve. Fourteen. PM.” No year. Furthermore, while knowing the year would allow the machine to compute whether February has 28 days or 29 (assuming the machine knew or cared about the month and date), it makes no difference to the progression of days of the week.

So why does it prompt for the year? How will it know when it should change to 2008?

Upon reflection I can think of only one reason — an evil one — for requiring the user to enter the year: programmed obsolescence. When it gets to be perhaps 2009 — 731 days, more or less, after the user sets it to 2007 (that’s 365×2 days, plus one for leap year) — maybe AT&T plans to have a new model out so my answering machine will just stop working and I’ll have to buy a new one.

Which begs the question: why, knowing or suspecting all these things, did I obediently set the machine’s year to 2007?


This morning I spotted this shutterbug.com link about commercial food photography on del.icio.us’s popular page, which was my cue that it’s time to tell my Zanetti story.

In the late seventies and early eighties my dad worked as a salesman for a bookbindery. One of his clients was The Creative Black Book, a celebrated annual directory of and for professionals in creative services such as filmmaking, copywriting, photography, etc. The Black Book staff, led by founder Marty Goldstein, was young and hip, and their offices were fun to visit. My dad found plenty of reasons to pay “business calls” on this client that were really excuses to hang out and goof around. Goldstein and my dad became friends.

One of Goldstein’s friends-and-clients, in turn, was a photographer named Gerald Zanetti, who had a thriving commercial-food-photography studio in midtown Manhattan. Zanetti was also a technophile who in 1982 had one of the first small-office computers — the TRS-80 Model II — and no software to run on it.

Who you gonna call?

In 1982, Zanetti had a new computer but no way to use it, so he called me. The following year, my high-school friend Joelle bought herself a car but didn’t know how to drive, so she called me — letting me keep and use the car myself for many months in exchange for occasionally driving her places.

A tenuous connection between these two stories? Maybe. But today, Joelle’s husband runs the TRS-80 website hosting the Model II link you see in the main part of this story!

Enter yours truly. After a few years as a programming hobbyist I was ready to earn some afterschool money writing software. Through Marty Goldstein, my dad hooked me up with Gerry Zanetti and I had my first programming job.

(Holy hell, that makes this year the silver jubilee of my career.)

Zanetti loaded me up with a bunch of documentation about the Pickles and Trout version of CP/M, an operating system and programming environment that famously was almost the basis of Microsoft’s original DOS. Within a few weeks I was writing some simple office-management tools for Zanetti, but in the classic fashion of non-technical enthusiasts with money to burn, the specs for what he wanted were ever-shifting. Far from a consulting gig, my position at Zanetti’s studio became open-ended. When the school year finished I worked full-time over the summer. I became part of the Zanetti studio family, which included Zanetti, his wife, a rotating slate of assistants, and recurring visitors such as suppliers, food stylists, and other partners. My sister Suzanne did office work there on occasion too, and even I occasionally answered phones, stuffed envelopes, and ran errands.

As I toiled at the computer on a wide variety of projects — now programming, now data entry, now educating myself further in CP/M and later Xenix — all kinds of fascinating photgraphy-business stuff went on in the adjacent studio. I saw all the tricks of the trade described in the shutterbug.com article and then some: cereal arranged in a bowl of white glue, not milk; cigarette smoke blown across a cup of coffee to simulate steam; mashed potatoes for ice cream; and so on. For most jobs Zanetti used a professional home economist (a.k.a. food stylist) to prepare the food to visual perfection. I never knew a Whopper could look so good!

He was among the best in the business and lots of big-name clients came through the studio. Very often they purchased excessive amounts of food for the shoot and left it behind after the shoot was done. Many were the times I lugged a dozen steaks, a few hundred slices of American cheese, or a crate of Ronzoni spaghetti home on the subway.

Eventually I left New York to go to college but I continued doing bits of work for Zanetti from time to time via a character terminal and modem he supplied me with. He could never get enough of me writing software for him though he needed almost none of it, and his wife occasionally gave him a hard time about the expense of it. He was full of ideas for new computer projects, some quite ahead of their time — his blue-sky description was the first I ever heard, for instance, of an object-oriented photo-editing application.

Zanetti exhibited paternal forbearance while I learned my craft on his dime. I was exceptionally lucky to have such a willing patron and would not be where I am today without his support. Thanks, Gerry.

What brings you here?

Herewith, a selection of search-engine queries that resulted in hits on this blog, according to my server logs.

watch neighbor undress; exploratorium; fizzies pulled; thailand’s greatest hits; what happened to fizzies tablets; what are the three kinds of mammals; violet incredible pez; watching neighbor undress; “new rabbi”; Rosh Hashanah; “the federation trading post”; persian candy floss recipes; “citric acid” science pop candy; supine lady; lesbian sex; ursula.sex; how to explain the theory of crystallization to third graders; evil cats; i feel like crying; Comcast Removes West Coast Feeds; boycott disney & abc path to 911; raiders-of-the-lost-ark Pirates-of-the-caribbean; “name the moon” greg; Reality an space-time; “francis heaney”; “smut shack”; squeamish cure; doggie style sex; webby awards; amy linker; cynthia nixon; quarks tangles; “mill valley pediatrics”; “dildo with suction”; proposition moveon endorsements; hypothermia kim; steve elliot bdsm; “instant soda”; trish gee wordpress; song meanings splashdown; “yours yours yours”; fligth to mars; “lesbian vampire fiction”; “needed a diaper”; disney fingerprints; fizzie drink discs; ben kenobi obgyn; linux backup s3; melissa kaplan sings; “adam stoller”; splashdown catalogue; “santa claus ain’t”; vote to boycott abc disney path to 911; joe costanzo; doggie style sex positions illustrated; Fizzies drink tablet recipe; “jack mccoy”; charteris; tune out, turn off; incremental jungledisk; “no fireflies” long island 2006; vampire lesbian; Thai Pilot; boisterous laugh audios; simpsons ulysses; sephardic pirates kritzler; backup osx hardlink incremental; karma slave karaoke; voyager pale blue dot send back the image; comcast digital artifacts; Con Edison; What year did the sitcom premiere I dream of jeannie; Recently got digital cable still receiving all premium channels; joseph costanzo, jr.; Superman reversing time; three kinds of meat; video koyaanisqatsi koyaanisqatsi; chabad palo alto; 9/11 personal; “Calculatrivia”; viscera at&t.

Curiosity killed the camera

Earlier today a co-worker sent a link to a pictorial of a camera being disassembled, the point being how fantastically complicated it is compared to the device we make. It reminded me of this story:

A couple of years ago, the hardware guru at work, Sue, let me sit at her workbench and use her tools to try to repair my Canon Powershot digital camera, a midrange point-and-shoot model. On powerup, a tiny servomotor was supposed to telescope the lens barrel out of the camera body (just as hundreds of models do). Mine had stopped operating smoothly. The barrel, or the protective shutter in front of the lens, would get stuck halfway through an open or close cycle. I’d already gotten an estimate on a professional repair that was prohibitively long and expensive. Might as well give it a try myself and, at worst, buy a new camera.

I thought that by disassembling it as far as the lens barrel, I might be able to dislodge any grit or whatever was blocking its smooth operation, provided it wasn’t actually an electrical problem in the servo or anything else. I took out the batteries, then started removing screws and laying them carefully on the workbench in a way that might allow me to remember how they were all supposed to be put back together. But I was only able to remove tiny bits of camera at a time. Dozens of removed parts later it still looked pretty much like a camera. It was like the dance of the seven veils.

Before long there were so many screws and bits of camera shell and buttons and retainer rings and spacers on the workbench that it was clear it would never all go back together. But I pressed on anyway out of stubbornness and curiosity. Until the camera blew up in my hands.

Yes, I had forgotten about the giant capacitor that powers the flash. It discharged painfully into my fingertips with a loud BANG, a tiny shower of sparks, and of course the magic blue smoke. And that was the end of that. Now a little pissed, I spent another minute or two manhandling the (now slightly charred) camera just to get a glimpse inside the lens barrel by any means necessary. When I finally ripped it apart enough, there appeared to be nothing obvious I could have done anyway to fix it. I swept all the pieces into the trash as the phrase, “No user-serviceable parts” repeated over and over in my head.

Ibid update

Ibid (“incremental backups to infinite disk”) is a tool I’ve written for backing up files. When I first posted it (read about it here, download it here) it was at release number 18. Now it’s at release 24. Here are the changes between then and now.

  1. If the same pathname is encountered a second time in one session, don’t treat it as a hardlink to the first time.
  2. Restructure the ~/.ibid directory and add an extra level to target pathnames when many sessions begin to accumulate.
  3. If the old session file was compressed, compress the new session file the same way.
  4. For determining whether a file is a mere rename of one already archived, consider its size in addition to its device, inode, and modtime.
  5. When running with –ensure, flag (in the session record) the files that have been verified as present in the archive, so that future sessions won’t have to reverify the same files.
  6. Remove an extraneous newline from “ibid -D” output.

Still to do: a proper home page for ibid.