Sep 04 1 2004


At Hypertext '04, I spent a delightful day touring Napa with Hugh Davis, who knows more than somewhat about wine. Hugh had brought a handy little bookshelf of well-written wine books for plane reading; as soon as I got home, I sent off for a copy of each.

Jancis Robinson's How To Taste suggests that the best way to learn what's what is simply to compare. If you open two or three bottles at once, taste each, and enjoy them all for a few days, she suggests you'll learn a lot more than by tasting one at a time. Unlike lots of wine writing, Robinson's prose is clean and sensible, and the book is filled with fun and tasty exercises.

In The New Yorker, Gopnick has a gem of a review, occassioned by William Echikson's Noble Rot. Interestingly, Calvin Trillin has a long New Yorker piece on wine as well, to which Gopnick alludes. This gives us a wonderful opportunity to compare Gopnick and Trillin at the same table.

Dylan Kinnett is uncomfortable with weblogs whose mission is primarily to sell the author's books.

What I really want to read is something literary: notes on reading, writing, etc. and not something commercial.

But this can't be right. What writer is disinterested in their own work and their own audience? A dead writer! But what interests us in literary weblogs is that they aren't dead at all. They unfold every day.

The problem, really, is not that the weblogs are commercial, but that they can become untrue -- dishonest in their own terms. This, in the end, is also the core of the Cluetrain gambit: people don't mind your extolling what you really believe, but they hate it when you say things about your product that everyone knows are untrue.

I like to hear a writer explain why they wrote the book, and why it's important for me to read it. I'm much less happy about reading suborned reviews from the writer's friends and family, or snarky reviews of the author's rivals dumped onto amazon by anonymous "critics".

Doug Miller is planning to attend the upcoming Tinderbox Weekend in San Francisco. He's got a new topical weblog in Tinderbox, to supplement his main weblog (and his main use for Tinderbox in his real-estate business)

Tinderbox lends itself to special-purpose weblogs, especially when its unique tools for experimental information architecture can prove useful.

Also, nice words about Tinderbox for the Engelbart target of augmenting human abilities from Oliver Wrede, too.

Today, I wanted to take a short time-out and look over a bunch of editoral ideas for TEKKA. We're just finishing TEKKA 6, and TEKKA issue 7 is already looking pretty solid. But what about TEKKA 10, say? What should we be talking about next summer, and whom should we want to talk about it?

This is classic whiteboard territory, except that we all know that no whiteboard is ever going to be big enough. That's one of the nice things about Tinderbox -- it gives you an enormouse, scrolling whiteboard.

A Very Big Whiteboard
  • One of the nicest things about Tinderbox for this task is that you can pick up and move things. Real whiteboards get smudgy partial erasures when you try to reorganize.
  • Another nice thing: you can open each box and get a nice text window for notes. When an idea originated in an email note, I simply dropped the email inside the box -- just in case I want to refer to it later.
  • There are 87 things in this map — a lot for a whiteboard, but a piece of cake for Tinderbox.
  • I can share the real thing with the TEKKA staff, even though they're all over the world. And I can share the picture of the whiteboard with you, right here on this weblog. Sharing is fast.
  • I'm using a few bits of Tinderbox magic to lend a hand here. You can't see them at this distance, but ideas about fiction have thin red borders, while ideas about reviews have thin blue borders. Near the top, I've got agents that automatically collect articles that are pencilled in for upcoming issues. Links connect tasks (darker red boxes) to article ideas (lighter boxes), so things stay connected and information doesn't get duplicated.

Not bad for a side-effect of a brainstorming session.

Aug 04 25 2004

A year ago

Everything in this post is about hypertext. It has no links....

I don't often single out a year ago notes, but this is a good one -- and one that wasn't much discussed at the time. One of the nice things about Tinderbox is that features like this are easy to build yourself, A Year Ago is actually a pair of agents; one finds all the notes written within a three or four-day window a year back, and the second selects the note that seems most substantial from the shortlist.

Eastgate's running a terrific sale on Tinderbox right now.

Imagine what would happen if George Washington came back and ran against George W. Bush. Did he really deserve all those honors? I bet he wasn't even there at Horseheads! And look at how he led his men into terrible danger — on Christmas Eve! — without proper rations or equipment! This man recklessly endangered his comrades: he is Unfit To Command!

Louis Menand has a predictably fine review of the depressing evidence that almost no voters have a coherent idea of what they are voting for, or why.

Aug 04 21 2004

Perfect Blue

Satoshi Kon's anime feature Perfect Blue goes right to the top of the Hypertext Film Festival Hall of Fame (Memento, Sliding Doors, Mulholland Drive, Minority Report, Run Lola Run, Waking Life, Nashville).

Perfect Blue
Copyright Rex Entertainment Ltd

It's the story of a Japanese pop idol who is coaxed into becoming an actress. She gets bit parts and gratuitous sex scenes, which she hates. Her old girl group does better than ever. She buys a Macintosh and discovers a bunch of fan sites -- including sites that follow the singing career she abandoned, sites that seem to leak out of some alternate reality where she made different choices. And then, she starts getting visits from her alternate self, the pop idol.

All this is just the beginning: before we're done, we're going to run through just about every tool in the metafictional workbench.

There's a certain Peter Pan impulse at work here: I won't grow up! It's politically treacherous, especially in a medium that's so often concerned with schoolgirls' underwear. Don't let it worry you; the screenplay, by Sadayuki Murai, may be all over the place but it's definitely not naive. In fact, Murai goes out of his way to show you all your greatest fears -- deus ex machina, it-was-all-a-dream, we're so artistic it doesn't have to make sense -- in a sort of metafictional Hitcock.

Perfect Blue is very conscious of itself. It's got buckets of allusions. And it's stylistically fascinating; because Satoshi Kon uses an animation style in which nothing moves unless it has to, you're always very conscious of the artifice. (Kon also loves wild camera work, like long , long dolly shots that would be a pain to shoot on film. Of course, this is animation, so long takes aren't a problem for the actors. It seems odd in an animated feature, but it's fun in its own way.)

Tinderbox on the big screen at ISEA, via Diane, via Jill. (Who is the speaker?)
Tinderbox Sighting

update: It's Ben Russell. Thanks, Peterme!

In 1995, the US accepted 292 refugees from Montserrat, where a series of volcanic eruptions causing havoc. They were admitted to the US with temporary protected status", so they could live and work in the US during the natural disaster. (Lots more people from Montserrat went to England, which used to own the island, or to other Commonwealth countries, or to other Carribean islands)

Now, the Department of Homeland Security is going to deport the volcano refugees because the volcano is not calming down. Because the disaster continues, the nice Republicans at Homeland Security have decided that the situation is not temporary, and so the "temporary protected status" should no longer be operative. (Thanks, BBC)

What's the matter with new media criticism?

Yes, there's some lively, thoughtful work. Rich Higgason's new study of Mary-Kim Arnold's Lust (presented last week at Hypertext '04) is quite fine. We've made some progress, I think, from the time just three years back when serious people seriously though that giving a cash prize to someone-or-other was all we needed.

Writing is hard. Reading is hard. You've got to work at it, and you've got to know what you're doing and how to do it.

Look at Adam Gopnik's The Big One, in the August 23rd New Yorker. (Look now: New Yorker links go stale) It's a masterful review of a slew of recent studies of World War I, informed by an even larger bundle of older histories. Gopnik offers strong opinions, strongly defended, but he also reads everything with sympathy and explains each position with care.

Two kinds of 'inevitablism' have long held sway as explanations for the deeper sources of the catastrophe. One, made famous by Lenin, and still cited by some historians on the left, is that the war was the certain consequence of imperial overstretch and colonial rivalry.... Of this hypothesis, nothing really remains. The German Weltpolitik, the new historians tell us, for the most part drew Germany away from the European heartland, into minor skirmishing on the periphery. The globalization of the world economy, in turn, which in the first decade of the last century had reached a peak to be equalled again only in our own time, depended on peace. The bankers and industrialists were the last people in Europe who wanted a war.

I've read none of the new work he discusses, but because I've read several of the older pieces and generally agree with his judgments I'm inclined to trust him on the new stuff.

I'm not especially interested in WWI. It's not a period I study. I'm certainly not shopping for books about it. I spent my dinner hour reading this review last night because it is so good, not because I wanted buying advice.

Most important, Gopnik's telling us about the Great War, about what it means now, and about what History means and inspires. Our leaders are just as foolish, our zealots just as ignorant, and our weapons just as terrible as those of 1914. This is a conversation worth having.

At the Hypertext '04 workshop on Spatial Hypertext, Cathy Marshall showed some preliminary data from a study of the way real people read The New Yorker -- how they turn the pages, how they use fingers to hold their place, how they move on the sofa, how they flip here and there. Reading is a lot more complicated than people think.

Then, look at Julian Küklich's review of First Person, Noah Wardrip-Fruin's and Pat Harrigan's anthology of game studies and other new media reflections. (I'm a contributor to First Person, but Küklich doesn't particularly savage me. In fact, he praises my contribution for its apparent lack of technological determinism. That's nice. He did forget to mention my confit de canard, though, which is often tasty and which is every bit as relevant to Greco and Bernstein's essay on Card Shark and Thespis as technological determinism.)

He takes Jill to task for reasons I find obscure. He has some questions about the Online Caroline, the work she studied -- something to do with Motte's approach to Nabokov. He'd have liked her to write about this; she wrote about something else. Brief essays seldom answer all the questions a specialist might ask about a work or a genre, and if you ask little essays to carry all your luggage, you're destined to spend a lot of your life watching your suitcases tumbling about.

Here's his ringing conclusion: the book's numerous writers have not solved all of his worries.

Some of them are academics and some of them are practitioners, but they all have one thing in common: they have no clue what they are dealing with. But, then, neither do we, the users of new media. It is reassuring to know that we are all in the same boat.

You know, Fussell and Keegan and Ferguson and Strachan don't solve that Great War, either. You could say (if you were in a sour mood) that, collectively, they don't have a clue. But sour moods are contagious: how much nicer, instead, to look broadly and generously for ideas, and then to weigh them with care and sympathy?

My point is not that Kücklick's review is bad (it's not), or that I disagree with it, or that I think you shouldn't read it. The problem is that it's about buying the book, not the subject. And, as a subtext, it's about whether the contributors should keep their jobs or be exiled from academe. The latter topic is unseemly, and the former is less than compelling.

  • Reading a long review to decide whether to spend $27.17 is a doubtful proposition
  • Lots of people who would normally find Kücklich's review and who might follow his points about critical theory are instructors; they can get a free copy from the publisher with a phone call.
  • Lots of people who will read this volume are the students of those instructors. Once the text is assigned, they're rather stuck.

Why don't we talk more about the things that really interest us -- enjoying new media, crafting fine software, saving the planet -- and less about buying decisions and tenure points?

Anders Fagerjord has been experimenting with Tinderbox for time management, and has a lovely, lightweight approach,

Lightweight Planning

Why do computer scientists acquiesce in publishing things we know aren't right?

Take, for example, a paper in the latest JODI describing The PlumbingXJ Approach for Fast Prototyping of Web Applications. The paper looks mildly interesting, although it ignores a decade of important work in hypertext structure. But lets look at the motivation (section 3):

The use of software engineering techniques is essential for the development of hypermedia applications (Murugesan et al. 1999)

If you read and review computer science papers, you'll see homespun homilies like this all the time. I'm not unsympathetic, The only problems is: we know this is untrue.

  • Lots of important and successful Web sites and hypermedia applications have been developed by people who have no background in software engineering at all.
  • If you write down a list of hypermedia pioneers, how many software engineers do you find? If you write down a list of topflight Web designers, how many software engineers do you find?
  • Software engineering is not the core competence of leading Web design agencies. Firms with lots of software engineering chops have not established much of a beachhead in Web design.
  • As the Agile/Extreme Programming movement has called into question whether software engineering is essential for developing software, is it prudent to claim that it is essential for developing hypermedia?
  • The footnote appears to offer a source of evidence to back the claim. It does not; it's entirely rhetorical.

In other words, the authors got carried away here, and the editors and peer reviewers didn't bother to check.

There are at least 38 different kinds of windows in Tinderbox. I know this, because there are exactly 38 classes that compute a WindowTitle. (Some windows, of course, have no title, or don't need to computer a title). Indulge me while I adduce some consequences:

  • Because it's obvious that each HypertextView needs a title, this was one of the oldest parts of the Tinderbox code. Walking down memory lane is sometimes pleasant, but this youthful indiscretion summons up no bittersweet smiles.
  • The old method was named MakeWindowTitle. In prose, a bad choice like this would have been caught by my editor, or it would by now be way back in the rear view mirror. In code, we've had to repeat the same awkward construction in more and more places
  • Because this is an old corner of Tinderbox, the calling convention was designed to reduce memory allocation which I then thought would be expensive. Wrong: it's not expensive anymore. Wrong twice: the design doesn't actually reduce memory allocation. It did make the code look clumsy, though.
  • Ripping out the old code and replacing it took all 112 minutes of Wonder Boys, which is a very fine film indeed.
  • I think there's a business case for cleaning up the annoying code smell surrounding this. But I'm not at all sure. I can rationalize it, of course. I'm the boss, it's the end of a long day, it's 1 AM, and why not fix up the clutter. But it's like office furniture: you may suspect good furniture helps people work better, but it's hard to prove it.

Incidentally, the development peekhole is back.

Aug 04 17 2004

Help Wanted

Eastgate needs a part-time web wrangler to lend a hand moving some of our sites to a new server. Help cope with certificates, DNS, and light Perp testing and debugging. We need help right away, over the next 2-3 weeks; this could lead to ongoing work. Email or ichat .

Aug 04 15 2004

Flying Time

What? Mid-August, already?

Dylan Kinnett is pondering summer's end, and with it the conclusion of his internship at Eastgate.

I never would have dreamed, when I took my first tentative steps into the realm of this exciting new kind of theory, that i would be able to get so close to the very topmost and prominent people and events in that field. Why, just this morning I had coffee with Elin while she related to me some of the news from the Hypertext Conference....

Read the whole thing if you think it might be fun to intern at Eastgate. Both Dylan and Jessica have been tremendous to have around this summer, and though we've got a lot done there is always lots more to do.

Internships don't have to be restricted to summer. Interested? Email me.

Aug 04 12 2004


Jim King is currently delivering a keynote here at Hypertext '04, discussing pdf as a format case study.

He's jumped right into the inside of the pdf file format. This is something we don't do very often, perhaps because it's so close to implementation and we like to take implementation as given. It's hard, in the context of a scientific conference, to talk convincingly about implementation: it's hard to know what it right, and what is simply what we did.

I find this especially interesting right now, because I'm very interested in the problem of planning architectures for Tinderbox and for Storyspace III that will facilitate research collaboration. One approach, clearly, is simply to make the file format sufficiently clear, flexible, and easy to read and to transform. Hmmmm....

Aug 04 11 2004

Tinderbox maps

At HT04, David Kolb gave a fascinating talk about the process of hypertext writing, complete with lots of fascinating Tinderbox maps. (Kolb turns off curved links, which gives all his maps a distinctive flavor)

Tinctoris displays its own map -- a colorful gem,

Tinderbox maps

After a long trek through the Santa Cruz redwood fog, we arrived at Hypertext '04.

While the rest of the hypertext research community treks up the hill for Doug Engelbart's opening keynote, we're watching Engelbart's famous Demo, which is often cited as the great computer demo of all time. There's always something new to see -- new bits of detailing (branches of argument appearing on the screen in counterpoint to the central issue Doug is addressing, clever little graphic hacks for ancient character displays) and new historical topics (the world debut of the mouse, the mouse cursor, and the auxillary keypad).

Aug 04 3 2004


If I'm absent, it's not just the vagaries of conenctions on the road, or neglect induced by the proximity of art and archaeology.

Prosperity, the Tinderbook, had been acting badly for two weeks before the trip. Before leaving, I decided I'd better get a new hard disk, and opted to have a pro come and install the thing.

Unfortunately, this particular professional thought that drag-copying a disk in Finder was a good way to proceed with replacing the system disk. That turns out not to be the case. I was at the office at 1am fixing the disk, with a taxi to the airport slated for 6.

And I'm still finding problems with permissions, everywhere. I'd have spent less time doing the whole thing myself, and saved a few hundred bucks.

Nothing at Pasquales' was quite as good as the chiles rellenos from the last trip. (Well, the lemonade was perfect, but that doesn't really count) But those chiles were hard to beat, and everything was very, very good.

SantaCafe was romantic as ever, even in the rain. Elegant little spring rolls with cactus and something from the onion family. A lovely rack of lamb. Can't beat it.

And the margaritas at Coyote Cafe are worth every penny.

Tasting Notes

Tonight (speaking of pennies) we're off to Geronimo, which is one of my favorites restaurants anywhere. And then, tomorrow, we saddle up the donkey and head back up the trail to San Francisco and then to Hypertext '04.

I think, right now, that Santa Fe might be the best place in the world for a Sunday painter to see a lot of interesting new art. The city is absolutely packed with wonderful galleries that range from tiny labors of love to massive enterprises. The variety is immense, and there's not an overwhelming layer of schlock. (Yes, there's plenty of cowboy art, but it's much better art than the Lighthouse Porn that dominates the Maine coast)

Santa Fe has a long tradition of modernism, and also a long tradition of representational art, going back to Mabel Dodge Luhann and the Taos School and to a postwar suspicion of pretentious intellectualism Back East. That works nicely for me, anyway, because I overdose quickly on Minimalism and can only take so much Conceptual Art at a time.

Here in Santa Fe, you can walk from an exquisite collection of Two Grey Hills rugs to a striking show of Bateman's wildlife art, from a gallery filled with Brassai and Ansel Adams to some Picasso and on to some nifty new bronze sculpture, not all of which is sentimental. Lots of very fine work in pastel, a few memorable watercolors, and lots of great figures and portraits.

Santa Fe