Mon Jan 01 2007 21:04 Cleanup: Today we did some cleanup and tried out the paper shredder I got Sumana for Christmas. Our paper trails are better-ordered now, and I also took the opportunity to finally make our wedding album.

This was a project that had been maturing for some time. The album is a shiny thing from Pottery Barn (people like to give us presents from Pottery Barn for some reason) and the pictures were from Sumana's sister. The two had been lying in proximity for many months, in the hopes that the album would absorb the pictures by osmosis. Today I decided that wasn't working and manually stuck the pictures in the album, if you can imagine such a primitive task. Now if you come to our house we have something to make you look at.

Mon Jan 01 2007 21:12 Retrogaming Times Monthly: There used to be an online magazine called Retrogaming Times. Its publishing schedule became erratic and it eventually stopped publishing. Some of the contributors started up a new magazine called Retrogaming Times Monthly. Since there's a publication frequency right in the name of the magazine, it can never go under! And the back issues are still fun, since the topic is not one that requires news hooks.

Tue Jan 02 2007 23:29: The Medieval Bestiary collates the natural-historical inaccuracies of medieval sources into a single post-Borges work. I'd like to see the Roman and Greek sources included as well, but this is fun.

The echeneis is a fish, half a foot in length, that clings to ships and delays their passage. When this fish attaches to a ship, even in the high winds of a storm the ship will not move, but seems to be rooted in the sea. The echeneis is found in the Indian Sea.

Includes bonus coverage of stones:

The stone called oyster produces pearls. At dawn the oyster opens and takes in the rays of the stars, the moon and the sun, and also swallows dew, and from these come the pearl. The agate stone is attracted to pearls and so can be used to find them.

Does not include hit dice listings for some reason.

Wed Jan 03 2007 15:57 Naming Names: A writer's exegesis of his character names. The SSA's baby name rankings really are an excellent tool. My noir story takes place around 2015, and I used it to work backwards towards probable names, with the satisfying result of exobiologists and bureaucrats named Cody and Madison and Amber.

Wed Jan 03 2007 16:05 Speaking of names: Reading these old computer-history books it occured to me that it's been a while since people named computers with big acronyms ending in -AC. You used to see this in science fiction as well (Multivac, EPICAC) but again, not so much these days. I think it's time for this naming technique to make a comebac. Apple could call its next computer the UNIMAC or something.

Sat Jan 06 2007 10:10:

"It's like you don't care!"
"I care!"
"It's like you don't care!"
"I've failed the care Turing Test!"

Sun Jan 07 2007 21:33: Rachel got me some Frommer's books for Christmas. There's one I can't find at the moment which is about day trips that don't require a car. That one's kind of talky, but NYC Free and Dirt Cheap is great, full of things I didn't know and things I kind of knew but never really planned to do, arranged on the general theme of cheapskatery.

Mon Jan 08 2007 21:14: Adam Parrish invented the Micropoet 200, inspired by the Eater of Meaning. It runs on an applet that applies Eater-like transforms to text, including a cool one that makes text degrade the way Nethack engravings do.

(3) Tue Jan 09 2007 21:13 What happened to cybernetics?: Cybernetics was pretty big in the 40s and 50s. People had all sorts of grandiose goals for it: using it to understand living and social systems, for instance. But it kind of died out; it's the one thing in Stanislaw Lem's fiction that seems dated.

I'm pretty sure cybernetics died out because it had a mechanical feel and didn't fit well with the simulation-based philosophy of the universal computer. But simulations need to be simulations of something, so why not feedback loops?

Wed Jan 10 2007 14:15 Call it something I ate: I just realized that Infinite Jest would make a good episode of House.

Thu Jan 11 2007 21:39 Yay Report 01/11: Yay:

• Royalty check!
• Finished my asteroid-mining story!

Un-yay:

• Book deadline at end of month
• I got a search request for a day in the life of a piece of gum. I can think of three ways that can go, and none of 'em are good.

(2) Fri Jan 12 2007 10:59 When to Stop: I'm learning more about Mount Rushmore than I ever cared to, which is kind of distressing as I already knew more about Mount Rushmore than I ever cared to. I chose it as a minor example for the REST book because it's a well-known place about which people can have drastically different things to say. But now I'm finding out things that are just ridiculous.

For instance, the original scope of the project included not just busts of the four presidents but sculptures of their torsos. It would also have included "a massive panel in the shape of the Louisiana Purchase commemorating in eight-foot-tall gilded letters the Declaration of Independence, U.S. Constitution, Louisiana Purchase, and seven other territorial acquisitions from Alaska to Texas to the Panama Canal Zone." And Washington would have been wearing one of those big foam "USA #1" hands. You know all those SUV ads that have the SUV climbing Mount Rushmore? They would have been moot because there already would have been an SUV carved into the top of the mountain.

I tried to go to the original source material on that (original source: artsy-fartsy art magazine) so as not to be passing around Wikipedia hearsay, but original source material was 1) behind a paywall that 2) pretended not to be a paywall for the Googlebot but 3) pretended not to be pretending not to be a paywall when I pretended to be the Googlebot. So I'm going to believe Wikipedia's characterization of the article just out of spite.

Fri Jan 12 2007 23:20: Here's the No Twinkie Database, a more general list of game design flaws than my platformer-specific list.

Sat Jan 13 2007 18:31: Sumana's (finally) reading the Baroque Cycle. Unfortunately it would seem as though the supplementary wiki has gone commercial.

(1) Sat Jan 13 2007 19:27 Signed, Sealed, Repeat: Went to the Met today to say goodbye to the lovely Chinese writing exhibit. Today I decided I would get to the bottom of a mystery that's been bothering me: why do the manuscripts and paintings on display have so many seal impressions on them? Once I noticed this it really started bothering me and I kept looking at the seal impressions instead of the manuscripts. Random manuscript, random painting.

I don't understand a word of Chinese, except as part of a much larger system, but I think I figured it out. In recent works you have one or two seal impressions to act as the artist's signature and/or studio stamp. (I learned this much from English descriptions.) In older works you have seal impressions in those places, but also impressions scattered throughout the document seemingly at random. I believe that those extra seals come from the previous owners of the work.

The impressions aren't part of the work because some of them got cropped when the work was remounted, and in some cases more stamps were added to the new mounting material. I found two pieces that had two impressions in common, but they were made a century apart by different people in different genres. That only makes sense if they were in the same collections for a while. Seals use all different styles of script, but if the seal is part of the work you shouldn't get extremely old pieces with modern-looking impressions on top of them.

Sometimes the impressions actually obscure a bit of the work, which is the sort of sloppy stamping I'd expect from a censor or customs inspector. But why would you need to approve or inspect the same piece over and over again, even over centuries? One manuscript had 90 seal impressions counting duplicates (fortunately, most of these were in blank space at the end), and one unfortunate masterpiece had 27 impressions all over it like misplaced kisses.

In general it seems like the sort of thing people did before certificates of authenticity and modern ideas of art preservation. As I mentioned, none of the contemporary pieces had an unreasonable number of seal impressions (though given the dadaist nature of some of those pieces I wouldn't be surprised to see one where they completely obliterate a text). However this is just my hypothesis, so I invite people who might actually know what is happening (John? Steve?) to set me straight.

Since a lot of the works on display were letters, I've got a far-fetched hypothesis that some of the seals might be postmarks. Doesn't seem too likely.

(1) Sun Jan 14 2007 21:55 Finite Jest: Today's Gutenberg text is The Jest Book, a prototypical 1001 More Jokes For Kids, which in the introduction mentions an ur-jester named Joe Miller. It presents Joe Miller as an archetype like John Bull or Jack Straw, albeit an archetype no one's ever heard of. I did a little research.

Joe Miller was a real person, an eighteenth-century "actor of farce". But his real claim to fame was having his name posthumously stuck on a book of recycled jokes: Joe Miller's JESTS: Or, the WITS VADE-MECUM. The real author, John Mottley, would have done better to put his own name on the book. But the need to capitalize on Joe Miller's late lamented hilarity took precedence over thematic consistency. Literary scholars agree that Mottley threw in the awesome phrase "Vade-mecum" as a sop to me, three hundred years later.

This uproarious success of this book caused "Joe Miller" to enter the lexicon in two different ways: as a name for any literary joke-boneyard (such as The Jest Book), and more enduringly as a term for the stale joke itself.

Sample totally unprovoked Irish joke from The Jest Book, in the style of 1001 More Jokes...:

A FELLOW on the quay, thinking to quiz a poor Irishman, asked him, "How do the potatoes eat now, Pat?" The Irish lad, who happened to have a shillalah in his hand, answered, "O! they eat very well, my jewel, would you like to taste the stalk?" and knocking the inquirer down, coolly walked off.

Good thing he had that shillalah, or there would have been no punchline!

If you crave entertainment more to the modern taste, here's a New Yorker article about the history of joke books.

Mon Jan 15 2007 17:34: At last, a cartoon that furthers my propaganda war against the Ice King. Includes John DiMaggio. via Ryan.

(1) Mon Jan 15 2007 21:28 My Favorite Mortal: We came up with this idea for a 60s sitcom of the fantastic, in the style of Bewitched or I Dream of Jeanie. One of the deities of the Greek pantheon comes down from Olympus and becomes espoused to a mortal, takes on a day job as an advertising executive. Not only is this the top job in 60s sitcoms but its theme of mortal-manipulation fits in well with the preexisting skill set. There are, I hardly need to point out, hijinks aplenty in this concept.

We didn't reach agreement on which deity it should be. If it's Zeus you can have a funny theme song about a swan/shower-of-gold encounter gone awry, but that's about all Zeus has going for him. What about Athena? Still working on the concept. There's no rush as it will be quite a while before time loops back on itself and the 1960s come along again.

(1) Tue Jan 16 2007 07:41 How Sergio Leone Stole Christmas: I just saw an abandoned Christmas tree rolling down the street like a tumbleweed.

Tue Jan 16 2007 22:24 Ancient Chinese secret, huh?: Seth asked "an actual Chinese art historian" about the seal impressions on Chinese paintings and got a detailed answer, which I summarize:

• Some of the seals are the artist's signature, but most of them are to "situate the painting in a social context".
• Some of the impressions are made by the people who mount the work. "Chinese paintings are remounted on average every 75 years... The mounter places a seal over the edge of the seam between the old and new paper, to show when the work was done and by whom." That explains the half-impressions.
• Collectors sometimes stamped the items in their collections, as I surmised.
• Sometimes people would write a poem on a painting in response, and (I assume) stamp the poem. This might be a joint effort with the artist (he does the painting and you write a poem about it) or it might come much later. "It reveals a particular habit of mind of the Chinese literati -- that the proper response to an aesthetic experience was poetry."
• "Most important paintings made before 1700 were eventually collected by the Qianlong emperor, so they tend to all have the same distinctive big seals on them." The Qianlong emperor was also big on writing on other peoples' paintings. "Many of us think that if he were going to inscribe so many of the paintings he collected, it's a pity he wasn't either a better poet or a better calligrapher." Zing!

  This painting was given as an example. Analysis:

  The inscription on the left is by the painter, Zhao Mengfu. The one in the middle is probably by the Qianlong emperor... But also it was Qianlong who put those two seal impressions right in the middle of the painting -- disrupting the main axis of the composition, which is the gaze between the sheep and the goat. So one thing to remember is that not everyone in classical China would have approved of the way the Qianlong emperor deployed his seals!

• The practice doesn't apply to contemporary Chinese art, but "it does continue with traditional-style paintings". The artist will stamp the painting, and you can stamp it too if you like, to prove you owned it.
• Museums often do a similar thing, labelling pieces with catalog numbers and the like, usually on the back. I had considered this on one of my previous musings, but forgotten about it when writing the previous entry.

In general, metadata adds to the history of the object, and the mere act of adding metadata doesn't damage the object. But you need to be careful about how you add the metadata, because poorly-applied metadata will damage the object. The specific example given was a very Seth-ish example about what kind of glue you should use when affixing bookplates to books, but hopefully there was a subtext that you shouldn't stamp right between the sheep and the goat.

Very informative! Thanks, DrK.

(4) Wed Jan 17 2007 08:53: I'm a Jolt Awards finalist!

Thu Jan 18 2007 20:18: This is even cooler than the Mandelbrot generator. Via clickolinko I found this excellent function which you shouldn't click just yet because I've got a more entertaining presentation of it. You're supposed to graph the function

1/2 < |_mod(|_y/17_| 2^{-17|_x_|-mod(|_y_|,17)},2)_|

over a certain range. Here's a Python program that does it:

#!/usr/bin/python
from dmath import getcontext, Decimal, floor
getcontext().prec = 550

# Tweak these to get a larger or smaller-scale graph.
X_STEP = Decimal("1")
Y_STEP = X_STEP

# Don't mess with these.
X_MIN = Decimal(0)
X_MAX = X_MIN + 105
Y_MIN = Decimal("96093937991895888497167296212785275471500433" +
                "96601293066515055192717028023952664246896428" +
                "42174350718121267153782770623355993237280874" +
                "14430789132596394133772348785773574982392662" +
                "97155171737169951652328905382216124032388558" +
                "66184013235585136048828693337902491454229288" +
                "66708109618449609170518345406782773155170540" +
                "53816273809676025656250169814820834187831638" +
                "49115590225610003652351370343874461848378737" +
                "23819822484986346503315941005497470059313833" +
                "92264972494617515457283667023697454610146559" +
                "97933798537483143786841806593422227898388722" +
                "980000748404719")
Y_MAX = Y_MIN + 16

fl = floor
# Performance speedup
FLOORS_DIV_17 = {}
def fl17(n):
  "A memoized floor(n/17) function."
  f = FLOORS_DIV_17.get(n)
  if not f:
    f = fl(n/17) 
    FLOORS_DIV_17[n] = f
  return f    

half = Decimal(".5")
def f(x,y):
  return half < fl((((fl17(y))*(2**((-17*fl(x)) - (fl(y) % 17)))) % 2))

y = Y_MIN
while y <= Y_MAX:
  x = X_MAX
  while x >= X_MIN:
    if f(x,y):
      print "*",
    else:
      print " ",
    x = x - X_STEP
  print
  y = y + Y_STEP

It uses the dmath library so I can call the floor function on an arbitrary-precision decimal. Anyway, you can run that program yourself or you can see my premade ASCII graph of the function If you graph the function over other ranges you can see other patterns, but none so striking.

Lots more interesting empirical math in this paper.

Fri Jan 19 2007 18:18 Deerstalkerstalker: I had an aggravating flu the past couple of days so I was in bed reading a big book of Sherlock Holmes stories. This was not a complete Sherlock Holmes, like the two-volume set I read about fifteen years ago when I had the flue, but it was all the stuff that was serialized in Strand and illustrated by Sidney Paget. These illustrations are interesting because I'd read from various sources that they, not the text of the stories, were responsible for creating the modern image of Sherlock Holmes.

This is true as far as it goes. Holmes is depicted throughout as being gaunt and angular though the text never goes further than thin and stringy. What I'm not so sure about is the deerstalker cap thing. It's not as clear-cut as "Paget did it." It looks like Paget introduced the deerstalker as an innocuous one-off, and then fans took it and spun it out of control, the way you get entire franchise novels nowadays based on an offhand remark in the canon.

We begin the hat search. The fourth story in The Adventures of Sherlock Holmes anthology is "The Boscombe Valley Mystery", and the first picture shows Holmes and Watson travelling in a train. Holmes is wearing the deerstalker (text: "close-fitting cloth cap") and the cloak, which is apparently an Inverness cloak (text: "long grey travelling cloak"). Very iconic. But they never show up again til the first story in Memoirs of Sherlock Holmes, "The Adventure of 'Silver Blaze'" (visible here. Again, the drawing shows Holmes and Watson traveling in a train, and he's wearing the hat (text: "eager face framed in his earflapped travelling cap") and cloak. It never shows up again for the rest of the book. This seems to indicate that the deerstalker and cloak are some kind of specialized outfit for traveling by train.

Doyle kills Holmes off at the end of Memoirs, published in 1893. Next is the novel The Hound of the Baskervilles, serialized in Strand in 1901-1902. Again Sidney Paget does the illustrations and again Holmes never wears the deerstalker, despite spending most of the book in the country, a faraway place accessible only by train.

Then, suddenly, in 1903, a new set of short stories: The Return of Sherlock Holmes. And suddenly the deerstalker is everywhere! He wears it (with a trenchcoat) in the first story "TAot Empty House", in which not only does he not board a train, he never goes any further afield than next door! The deerstalker shows up again in "TAot Dancing Men" (where Holmes and Watson are about to board a train), "TAot Solitary Cyclist", "TAot Priory School", and "TAo Black Peter". In none of those is Holmes wearing the famous cloak, and in the last three train travel is implied in the text but not shown explicitly. Note that the hat is not completely correlated with train travel: in "TAo Abbey Grange", for instance, Holmes is depicted in a train carriage, but wearing a jacket and top hat.

What happened? It looks like the deerstalker became part of Holmes folklore between the time of Holmes's death and retconrrection, to the extent that, as the world forgot the Victorian taboo on wearing deerstalker caps when not actually inside the train car, Paget himself felt pressure to draw the cap even when Holmes was nowhere near a train. The illustrations for Pursuit of the House-Boat, published in 1897, shows Holmes in what appears to be a Stetson, so I think it was after that.

That's cutting it pretty close, but The Definitive Holmes gives the date as 1899, when William Gillette took to the stage with a hit Holmes play. He wore the deerstalker on stage, probably based on the drawing from Boscombe or "Silver Blaze", and that sealed it. Proof of Gillette's mythmaking power is that he also introduced the curved pipe.

This doesn't explain the lack of hat in the illustrations for Hound, nor these 1892-1893 illustrations that show the hat prominently despite the absence of any train. Of course, if one person can take the wardrobe in one drawing out of context, so can another person.

I am now officially tired of talking about the semiotics of hats.

Sat Jan 20 2007 19:10: I'm having a lot of fun searching through the New York Public Library Digital Gallery. See, for instance, the Frank E. Buttolph collection of turn-of-the-century menus. Hey, and look at those NYPL stamps all over everything.

(2) Sat Jan 20 2007 20:09: Another thing I did while I was sick was watch a bunch of Space Ghost. Not Coast to Coast, but the original Hanna-Barbera animations. Usually sentences of that form are intended to imply that the "original" is superior, but sentences of that form don't usually have "Hanna-Barbera" after "original". Space Ghost is as bad as every other Hanna-Barbera animation, and the Cartoon Planet recontextualizations, despite being often puerile and childish to the point that Sumana won't watch them with me, are greatly superior by any measure.

This makes catching up on the Space Ghost canon a good way to use time when you're sick but can't go to sleep: even when you feel like you can't do anything you're up to the task of comprehending a Hanna-Barbera cartoon. The stories are dull, the animation is yabba-dabba-lame, and the orchestration is almost always this ponderous horn section stuff. On the plus side, the art itself is good, the guest villains are often excellent, and sometimes the horn section gives way to '60s jazz.

It's pretty clear that the scripts were written in total disjunction from the animation, because everything that happens is both shown and clearly narrated. It's like the script to a radio show with no foley. Made up example: "I'll open the door. What's that coming through the door? Some kind of... monster! It's got me! I'll just... use my power bands!" Actual quote: "I've got to stall for time!" Truer words were never said.

In the first episode, Space Ghost and his pals go to another planet and just complain about everything, like stereotypical American tourists. Hey, people live on this planet! Or did, until you blew it up. "I can only pray that these poor creatures will find a home." My heart bleeds for your concern, Space Ghost. How about not blowing up the home they already have?

I only remember one funny joke. Villain: "Behind that door is a super-mechanical creature who needs only one thing--the brain of Space Ghost!" Space Ghost: "Oh, that routine!" Lots of unintentional jokes, such as Tansit's reassuring: "No threat, Space Ghost. Just a demonstration of my newest weapons of destruction."

It was fun to see Zorak and Brak. Zorak sounds just like in C2C, but Brak has a kind of German accent. Like I said, the guest villains are the best part. Sometimes you get a Law and Order type intro where the villain finishes off some unfortunate cargo ship and then Space Ghost comes over to investigate.

Mon Jan 22 2007 16:18: This Star Wars theory makes a fair bit of sense, but I'd just like to point out that the only way to make the whole mess hold together is to give all the connecting knowledge to the two characters who never speak English.

(3) Mon Jan 22 2007 19:39: It really is nice to watch snow fall. People aren't kidding.

Wed Jan 24 2007 18:29: Wow, I didn't know there was a print version of Diesel Sweeties. It's like an alternate universe where all the jokes are toned down for a mass audience.

Wed Jan 24 2007 23:06 This Person Doesn't Stop Search Requests: A while ago. Today: 'one day in the life of bubble-gum' 'story'. If I get that search request one more time I will write a "day in the life of a piece of gum" story. This doesn't sound like a threat but it is.

Wed Jan 24 2007 23:53: I'm working hard on finishing the first draft of the REST book, but as part of my old-stuff-cleanup project, I've also put up some of my better stories from many years ago on segfault.org.

(1) Thu Jan 25 2007 22:00: I wrote about Atom and the Publishing Protocol thereof all day and I'm tired, so instead of real content let me present an addition to the page I put up yesterday. The list of my good Segfault stories now also has links to my three sons Be Dope stories, which were mellower than my generally low-pH Segfault writing.

Fri Jan 26 2007 00:14: I just got 419 spam from Flammy Babyface.

(2) Fri Jan 26 2007 11:21: This breakfast cereal has "the goodness of soy". What if I want the evil of soy?

Sat Jan 27 2007 21:31: Joe Mahoney reminds me that he did his own Segfault retrospective about a year ago.

Sun Jan 28 2007 19:12: Hyperion. Bop be ba de ba.

Mon Jan 29 2007 20:58 Story Titles:

• Twenty Twenty
• Nothing But Troubles
• User Stories

Update: this was supposed to go into my private notebook, but oh well. Feel free to take my lousy story titles.

(1) Tue Jan 30 2007 11:07 The Sparrow: Man. Not as good as Fiasco, but more psychically traumatic, so it earns a high place in the "first contact goes catastrophically wrong" subgenre. After a long and excellent buildup the ending lacked a lot of plausibility, but I admire a writer who's willing to develop well-rounded characters and then just kill them off (not a spoiler). The timeline jumped around a lot but it never got confusing except once when there was a flashback inside a flashback. Is the sequel good?

Tue Jan 30 2007 19:17: Wrote about HTTP caching today. The draft of the book is now basically done, except for Sam's chapter and the contributed section on Django. Caching was difficult to write about because it's a huge topic that I don't really want to cover in detail. I ended up giving just a couple recipes a server can use to say "cache" or "don't cache".

Wed Jan 31 2007 20:03: Rachel has a phone interview coming up and wants your advice.