Arnold Kling reports on an interesting study of artificial weblog communities. His results suggest that weblogs are valuable because they filter and distribute information efficiently. But who will reward the writers for the value they create. Kling offers an interesting speculation:

"No individual blog is the source of that value. I believe that this collective nature of the benefit of the blogging system is what makes it particularly difficult for an individualized economic model to be successful. Thus, I believe that neither the donation model nor the advertising model will prove to be viable (there may be some transitory exceptions). Instead, I think that payment mechanisms that reward collections of bloggers hold more promise for the long run."

I believe this is the first serious recognition of the economic potential of weblog cluster.

Jul 02 1 2002

Boxes and Arrows

John Schull is dreaming about a weblog tool that looks a lot like Tinderbox. His storyboards have some interesting features that might well be worth a try.

Jun 02 29 2002

Summer Taste

Go to Home Depot. Find a nice hunk of cedar. Ask the nice fellow to lop off about 16" for you. Hop over to Turner's Fisheries. Get a nice 14" hunk of Atlantic Salmon filet. Skin on, thanks. Rush it home, marinate a bit in vermouth, ginger, lime, and pepper.

Light the Weber grill (hardwood charcoal, please). While it gets going, have a bottle of a nice California Chardonnay.

Drop the plank on the coals. It will char nicely in a surprisingly short time, Remove it; put it on top of the grill, and put the salmon on the plank. Skin down, thanks.

Cook it for about 12 minutes. The smoke rising from the board will be quite dramatic. Enjoy. Top with diced orange and lime, seasoned with maple syrup, cayenne, and a little more cumin than you might expect. Slice and serve from the plank, with a big bottle of perry .

Christina Wodtke (whose sidebar Victor Lombardi rightly praises for its exemplary elegance) gives Jakob Nielsen a pasting over his study of Usability Guideline Compliance.

And "International Websites lag behind"... he looked at six! Six sites? What if he looked at the wrong six sites? It's like saying I looked at 15 American women and 6 foreign women and American woman are much better looking. If someone said that to you, you'd wonder what women he was looking at... and what women he missed.

The irony is that Nielsen taught the field how important statistical significance really is. His 1989 paper on "Hypertext Usability: The Matters That Matter" looked at all the usability work done on the early systems and tossed out all the insignificant results. The discard pile was awesome. Nielsen knows methodology and he knows statistics; why is he doing this?

One reason might be that his audience changed. In the 80's, Nielsen was a scholar and his audience was the research community; they needed rigor and precision. Now, Nielsen's audience is top management; perhaps doing things right frightens them.

Information architect Victor Lombardi writes about what happened when he made a mistake in a Tinderbox template.

This morning in my Inbox was a message from Mark with the corrected syntax. This afternoon, before I had a chance to change my code, was a new version of Tinderbox that made my syntax valid. That's service!

Agile development lets Tinderbox move swiftly, which sometimes lets us do "impossible" things like this.

The downside is that agile development doesn't fit well with the traditional review model. If you review an agile program in March, by July you may find all the issues you wrote about have been fixed (and no doubt a bunch of new issues have arisen). Strange old world.

Palladium is Microsoft's proposal for securing computers by encrypting everything at the hardware level, locking everything down, and (apparently) keeping the keys in Redmond. Cringely thinks its a lousy idea.

Speculation: the accounting crisis is going to get much worse, and integrity is about to be rediscovered as a virtue. The Republicans think people want security -- more surveillance, more arrests, more patrols -- when they want honesty -- honest effort, honest dedication, and candid accountability. Trust is going to top the charts; Microsoft's (again) getting off on the wrong foot. Palladium is so 90's.

MoveableType, a server-side weblog tool, has added a new Web service called TrackBack. It's novel, and very interesting. Thanks, Aaron Swartz!

Jun 02 26 2002

Wiki Way

One of the most surprising aspects of The Wiki Way (by Bo Leuf and Ward Cunnigham) is its emphasis on Wiki as a personal notebook.

Wiki makes a great tool for storing notes, but it's not really a wonderful tool for making them. You need to write in a browser, which is a nuisance -- and, for personal use, Wiki's write-anywhere ubiquity doesn't do you much good. Tinderbox is probably a better answer. (In fairness, Wiki runs on lots more platforms, and Tinderbox didn't exist when the book appeared last year)

We now have round-trip Wiki editing running in a beta Tinderbox environment. You download a wiki and drop it into Tinderbox, edit what you like, and then export the changes. Mirror the changed files, and you're set. The prototype Tinderbox follows WikiStyleLinks, so you have Wiki links as well as conventional Tinderbox links. (More in Books...)

I never learned to touch type properly; I improvised a fast hunt-and-peck method in grade school and have never managed to get my touch typing speed to exceed my hunt-and-peck.

Recently, I thought I'd give it another shot, and even bought Ten Thumbs for a little exta push. Unfortunately, despite the awards and the reviews, Runtime Resolution Ltd. simply got it wrong.

What you need most from a computer typing tutor is a combination of practice and pedagogy; you need exercises to improve your skill, and you need intelligent oversight to keep you from picking up dreadful habits or being swamped, or bored, because your progress is faster or slower than the program expects. That means the program needs to act like its paying attention; it should seem to be pacing you curriculum intelligently.

Ten Thumbs might be pacing me intelligently, but it constantly undermines its authority with silly, over-elaborate, and pointless illustrations and animations. There's this cartoon Viking fellow all over the place. Why? Does he convince me that the program knows what its doing? He does not. He convinces me that the product manager thought a cute Viking would sell the program to people who would inflict it on students, despite the evident waste of development hours and talent.

Hawkens' View From The Heart has some tales of critical care nursing that remind me of my father's stories about Cook County Hospital in the good old days.

Back to work. Hospital full. No beds, few nurses, and the usual fun and games that make critical care nursing the glamorous and fashionable profession portrayed in both song and story.

And it's June, which means the new residents in the medicine and nursing programs are on their way. Time to make sure my ibuprofen and famotidine bottles are full.

In a weblog post that could easily have more impact than many conference papers, Jill Walker raises the question of focalisers in weblogs, games, email sagas, and other digital narratives. The focaliser of a story is the person from whose point of view the story is told. The focaliser and the narrator (the person supposedly telling the story) might be two different people entirely.

It's a really useful distinction, obvious once you think about it, buried in Genette's Narratology.

Jill gives up too soon when she encounters tricky cases like OnlineCaroline, where Caroline is obviously a focalizer but the reader herself is recruited to become a focalizer as well. Borders are interesting places; the utility of theory is that it takes case that seem "weird" or "experimental" and shows us why they're complex. Improv theater, role playing games, theater in the round, hyperdrama, Thespian hypertext: they're all playing with the place of the focalizer in drama. That doesn't mean the theory failed, it means theory succeeded in explaining why these cases are unusual and unsettling.

An interesting followup question: what happens when the focalizer isn't the protagonist? I believe the identification of the reader with the hero-protagonist is the main cause of the artistic failure of computer games to date, leading inevitably to "My Friend Hamlet" and the adolescent trivialization of fate that afflicts even the most artistically ambitious games.

Jill says, "Maybe this is chapter three." I suspect it's going to be a widely read dissertation.

I've been working this week on building bridges between Tinderbox and tools like Moveable Type and Wiki.

For example, Wiki automatically builds generic links; whenever a Wiki sees a mixed-case (CamelCase words with InternalCapsLikeThis), it creates a hypertext link to a page with the same name. Generic links are a staple of open hypertext research tools. Generic can be both convenient and interesting, though they also introduce problems (like distinguishing when a link to "Washington" means "George", "Booker T.", "state of", or something else).

Just for an experiment, I added Wiki's primitive generic link facility to Tinderbox, creating an experimental version of Tinderbox that understands WikiLinks. This has two implications.

Personal Wikis used by individuals as a notebook or commonplace book are a hot topic in the Wiki world. When you come right down to it, though, Tinderbox is simply better at this task than Wiki. Adding support for Wiki links might make the advantage even clearer.

A Tinderbox-Wiki connection could be very interesting, both in terms of exploring the boundary between Wikis and weblogs, and for better cultivation of Wikis. Tinderbox, for example, could be a power tool for helping a Wiki editor with a major refactoring.

An emerging discussion of structure in Web logs provides a great example of the way weblogs are self-organizing hypertext.

First, Meg Hourihan observes that weblogs share genre conventions, and that these conventions -- posts, archives, chronology -- free weblogs from the limitations of pages. Then, Jeff Ward argues that Meg is describing a grammar of webologing, forming a pattern of connection through time that resists the simple, rigid hierarchy of the old homepage form. Now, Jill Walker notices that this echoes her work on narratology, and points out that the chronological argument mustn't be taken too far: "the chronology", she reminds us, "doesn't drown the separateness of the posts."

I, in my turn, question whether the separateness of posts matters much: nobody remembers Kaycee's separate posts, but everyone remembers her illness, her death, and then the discovery that she never lived at all. The gradual emergence of argument, connection, and narrative from the separate posts matters more than their separateness.

The most interesting aspect of the whole thing, though, is the gradual emergence of a pattern of connection that's much richer and more complex than an oral discussion.

Now and then, I answer the main phone line at Eastgate. (Everyone else at Eastgate tries to break me of this habit, reminding me how busy I am.) Often, people are surprised that I'm not a recording.

SpeedySnail reports the disorienting effect of the other side of this transaction: expecting a machine but finding, instead, you've reached a real person. "What's the point of all this dehumanizing technology if it isn't populated by dehumans?" (Thanks, Jill)

Jun 02 20 2002

Digital History

Edward L. Ayers opened Hypertext '02 with a stirring, inspiring discussion of the role of hypertext writing in the study of history. His wonderful Valley Of The Shadow site explores, in great depth, two matched counties in the era of the American Civil War. One county is at the Southern edge of the North, the other at the Northern edge of the South, and Ayers has gathered tons of information -- every available letter, diary, newspaper article, farm ledger, everything in sight -- and woven them into a wonderful historical narrative.

In an earlier essay on History in Hypertext, Ayers explains how it works:

In the Valley Project, for example, one often gets the sort of rush that Bush and Darnton describe, that shock that comes from the juxtaposition and connection of the unexpected. To search for a marriage record and come upon a death of a child, to see the heartbreaking ways in which soldiers’ letters use the word "home" arrayed one after another, to juxtapose confident predictions of victory from both sides--all partake of poetry as much as analysis, of association as well as explanation. Yet they tell us something worth knowing about these times and places. And they are created unpredictably, on the fly, in a continuous form of spontaneous poetry. A carefully constructed text could invoke those associations intentionally and to great effect, weaving together source and connection in new ways. (emphasis mine)

Even better: Ayer's 1999 paper on The Pasts and Futures of Digital History.

As anyone who has tried to write history knows, historians either have to hold our temporal breath while we look around or ignore the changing social landscape as we push ahead in time.

Historians might begin to take advantage of the new media, then, by trying to imagine forms of narrative on paper that convey the complexity we see in the digital archives, perhaps emulating writers of fiction in this regard even as we maintain our rigorous fidelity to the evidence. We might acknowledge more frankly the limitations of simple narrative or monographic abstraction. We might try writing in more self-conscious ways, manipulating point of view, chronology, and voice more than in our current practice. This need not be postmodern flight into chaos, but could rather be a more satisfying engagement with the complexity that we know characterized the past.

If you use MoveableType and can spare a few minutes to lend a hand to Team Tinderbox, I'd love to hear from you. Email Thanks.

Joel Spolsky's latest Strategy Letter does a nice job of explaining why religious wars miss the whole point of open source. Related products, like automobiles and gasoline, are "complementary"; their sales are related. If cars get cheaper, more gas gets sold. If gas prices go down, people tend to buy bigger cars. Computers and software are often complements.

A British scientist at Hypertext '02 was surprised to find out that Eastgate was behind Tinderbox. "I thought is was another clever way Apple had found to sell more Macs!", he said. At that very moment, one of his colleagues walked up and mentioned that he'd finally convinced their department to sanction a new Macintosh for Tinderbox.

Of course, Tinderbox for Windows is under development. But sometimes it's better to get started right away -- and computer pros need access to both platforms in any case.

Michael Rose says that he's really enjoying Tinderbox, and suggests that richer Applescript support might hold the key to interoperability.

The problem with scripting in this context is that the script needs to understand all the systems intimately. Subtle changes and bug fixes can easily break scripts, so each new release of any tool might break your existing scripts.

On the interesting experiments front, though, try opening Tinderbox files with your favorite XML editor. That's the advantage of Tinderbox's use of XML: no magic, just files that are easy to share with other software.

I've been experimenting with Unsanity Software's freeware Silk, a small extension to MacOS X 10.1.5 that adds text smoothing to a host of applications, including Tinderbox. It's a small extension to System Preferences, and seems both flexible and trouble-free.

At first, the apparennt fuzziness of anti-aliased type may be slightly unsettling, but in time it simply looks better. (Do make sure your font settings are large enough for you to read comfortably!)

Jun 02 18 2002

Round Trip Woes

At Hypertext '02, I had a long and fascinating discussion over Guiness, much noise, and the NBA finals with Frank Shipman and Gene Golovchinsky (at right), exploring issues of interoperating hypertexts. This is a hot topic in research right now, and also hot in the weblog world because Dave Winer's OPML is a distributed system and is built into Radio Userland.

A key problem is interchange: moving, for example, from OPML to Tinderbox and then to OPML again. Golovchinsky has great ideas for ensuring that two systems can pass data back and forth unharmed, even if they don't know much about each other. Shipman, on the other hand, argues that it's a hopeless proposition; different systems are different media, and round-trip interchange is as hard as going from a book to a movie and then adapting the movie into a book.

Liz Klastrup inquires for weblogs written by "professional" print writers.

This is a problematic category, because many notable writers are not, strictly speaking, professional: most writers earn the greater part of their income from teaching or practicing a profession rather than directly from their royalties. In a sense, the most professional writers are technical writers and journalists. But I think Klastrup is thinking of novelists.

Klastrup mentions Bruce Sterling's weblog at Infinite Matrix. Two print writers' weblogs grew out of book tours: Neil Gaiman and Shelley Jackson.

A hypertext tool for contemplating the organization of your work -- intended to serve as a sketchpad for concepts that must ultimately be expressed in a linear argument. "Design is the process of transforming words into space."

I'm writing, live, from the 2002 Spatial Hypertext Workshop in College Park.

Jim Rosenberg discussed the complexity of time and change in a spatial environment. Agents make the idea of time (and edit history) very complicated!

Frank Shipman described a host of interesting developments in VKB 1.1; he's thinking about geographically distributed spatial hypertexts.

Jun 02 8 2002


The custom of college reunions is, when you think about it, a very strange habit. I keep waiting for William Hurt (The Big Chill) to walk in and deliver the punch line: a long time ago we knew each other for a short while.

I went to Swarthmore a very long time ago, and the short while was remarkably intense. We didn't live well, but we worked too hard and loved too much and this weekend I've been sitting on lawns with old, old friends. Good times.

Jun 02 6 2002

Small Players

Dan Bricklin observes that small sites dominate the Web. Some big sites do get lots of traffic, but their impact is dwarfed by the vast number of small sites. (For an earlier view, see my HypertextNow look at the terrain Beyond The Portal, written years ago when the popualr wisdom held that portals were the only Internet properties of value)

My Name Is Captain, Captain, by Judd Morrissey and Lori Talley, is now available from Eastgate. It's a fascinating, intricate, and absorbing multimedia poem.

Zerospace is written by Tim Arch, a television worker and media theory student at RMIT.

Jun 02 5 2002

OS X: accounts

Why does MacOS X provide such conspicuous user accounts? Yes, it's based on Unix, but why do you need separate user accounts on a laptop?

One approach might be to create separate accounts for your roles and projects. I might have one account set up for writing, with its own dock, desktop, and fonts. Another account for code development would have its own setup. This might work in practice, or might make switching tasks inconvenient. What do people actually do?

Is there any ethnography on the subject?

Jun 02 3 2002

Good Cheese

People who want to poke fun at weblogs talk about the writing nobody cares about: "I had a cheese sandwich today." It's a hazard of bad weblogs.

On the other hand, a meal can lead to wonderful writing -- if you care about it, and if you can manage to explain why. Pyra/Blogger co-founder Meg Hourihan (who is coming back to the Mac) had a good meal: "It was the most amazing night of my life."

I'll see you in Maryland! I'll be at Hypertext '02; hope you will, too.

The Worcester Art Museum (which was recently host to a wonderful exhibit of Weegee photographs) has a huge, inkjet portrait mural by Julian Opie.

I find Opie's portraits and figures fascinating and haunting. The drawing is very concise -- essentially comic symbolism -- but so carefully realized that the effect seems just short of realism.

Jomosoft has a new Macintosh FTP client, osXigen. It's MacOS X only, but seems to have a nice way of mirroring Web sites cleanly while letting you specify exactly which files you want to share and which (templates, design notes, etc) you don't need on your server.

At the moment, I'm quite happy with the new Fetch. It was written at Dartmouth but its author, having won a lot of money from Regis, bought his work from the university and now supports it professionally.

There's been a lot of interest lately in "live blogging" at meetings and conferences, where writers in the audience can spread the word about news and opinion while the conferences is still going on.

Sharing information quickly is great, but sharing information deeply is even more important. That means having good tools for reflection, analysis, for sharing structure as well as reaction, facts as well as emotion. Jill Walker gives an interesting example of using Tinderbox to share notes on a recent conference in Trondheim, Norway on ethics and Internet research. You can read her weblog for an incisive introduction, and download her complete Tinderbox notes for more depth. (You can read the notes with the free Tinderbox demo; Tinderbox 1.1 runs on MacOS 8.6, 9, and MacOS X)

"The ability to change from a map view to an outline is a very intuitive way to revisualize information. Very flexible, very responsive.

"Ladies and gentleman, I think we have a winner here. Very, very cool." -- View From The Heart 

It's striking that, with all that's been written about Tinderbox, this is one of the first Tinderbox comments to dwell on one of it's nicest features: Tinderbox has lots of views, and you can use them all together. Switch instantly from maps to charts to outlines. Leave 'em open, or create new ones as needed.

Hypertext veterans are used to this, thanks to Storyspace. But it's a good old idea that's been too often neglected. Tinderbox was designed from the ground up to support this even more fully than Storyspace 1 did, and Storyspace 2 is already benefiting from the new technology. I'll be discussing this next week at Hypertext '02.

Evan Williams (Mr. Blogger) writes from the O'Reilly Emerging Tech Conference that

"I'm being seduced into the Mac camp (for a notebook, at least). I can't help it. They're everywhere. And beautiful. And everyone's in love with them. And the Unix command line and Java integration.... well, damn."

I've noticed for some time that computer professionals tend to get Macs with their own money, or Wintel machines as a free perk. Of course, as I've maintained for years, a lot of people need both.