Caveat Lector

Ye markup geeks, go read this paper. Right this minute. Cookies will wait. (Got there via xmlhack.)

It’s all about dealing with overlapping hierarchies in XML. It’s very cool. Good survey of work in the field. And it’s even moderately well-written.

Tell you what, though. I would not want to author the proposed XPath gizmo (which is something that I dimly envisioned while pondering the problem of reproducing print indexes via markup) in a text editor. No, sirree. Want a point-and-click tool. Likewise (and I am sure the authors of this article understand this), this system is bloody useless on a text that may change. I simply cannot imagine predicating a book-production system (paper or electronic) on it. Too easy to break the XPaths.

Oh, and while I’m thinking about it, how on earth is a system like this going to handle variant readings? All they’ve really done is pushed the problem of immutability one step down the tree: from the markup to the character data. Got news: character data ain’t necessarily immutable either.

Sigh. My kind of problem. Oh, well, at least someone’s working on it; it’s not like it’ll be ignored without me.

Ladies, gentlemen, fellow scholars, please! Can’t we find better rocks to split on than cookies?

Here. I hereby pledge a batch of cardamom cookies (my mom-in-law’s recipe) each to Jonathon and Burningbird. You may exchange for snickerdoodles if you wish. Not TimTams, but won’t they do?

My have-read-this-book score on the latest 100 best list is a reeking 38. Disgraceful. (I was pretty strict with myself: excerpts, abridged editions, or one story out of a collection didn’t count. Not that any of that would have raised me much past 40. Translations counted, of course. Not even my husband can read all those languages.) Oh, well. More stuff to look for at the library, never a bad thing.

Jonathon wants to know how I know he likes clackety keyboards.

Elementary, my dear Delacour: your use of red accents in your blog design demonstrates a drive to make noise in the world, while your careful watch over the ethical practices of the University of Blogaria pegs you as an individual who believes actions should have consequences. Your use of the Dishmatique shows that you appreciate technology that bends to your particular needs. Et voila! clackety keyboards.

I won’t tell you how long I spent coming up with that tissue of nonsense. The truth is that in your original post about keyboards, Jonathon, you linked to a description of an IBM keyboard whose most prominent feature is clackety keys.

No, the Kinesis does not have any special noises for enter keys or the spacebar. Sorry. (I bet they could be hacked, but I ain’t nothin’ but a Python tyro; messing with firmware is much beyond my abilities.)

Incidentally, today’s Slashdot has a discussion of an article from the BBC whose headline, at least, takes the rather inflammatory stance that ergo keyboards don’t do anything for RSIs.

As usual, the actual article’s position is more nuanced: RSIs have many causes, and keyboards won’t cure them all. Well, excuse the ugly noise, but DUH. Talk about your straw men. No, a keyboard will not fix you if your sole problem is your table. That hardly means that ergo keyboards are useless for everyone; for some of us, the keyboard is at least part of the problem.

I note also that the article didn’t go further afield than the Microsoft Elite in its hunt for ergo keyboards. No Kinesis. No Maltron. None of the chording keyboards. Nothing against the Microsoft Elite, but it ain’t all there is.

Forgot to mention something that Jonathon will find of interest. The Kinesis can be configured to emit clackety noises when keys are depressed, for those who like the audible feedback.

In fact, the out-of-the-box configuration clacketies. First thing I learned how to do was turn it off, before it drove me bananas. Still, for those who like that kind of thing, it’s there.

Yesterday was the first weekend-day with decent weather that we have had here in the Frozen North in quite a while. My husband finished and packaged up a manuscript for publication, and then we suited up and headed out toward the zoo.

We walked, as we generally do. We have to take a roundabout route, edging a golf course without a cut-through. Just as well; the necessary meandering takes us past what my husband calls Turtle Pond, a quiet bit of Lake Wingra where he used to eat lunch, back a few years when he worked in that area. We like to sit there. Sometimes there are turtles. Sometimes frogs. Sometimes fish. Once or twice a heron. Always water lilies and dragonflies.

It’s too early for the water lilies to bloom yet, and nary a turtle nose poked up from the surface. “If I were a turtle,” I said, “I would think it was a lovely day for sunning myself.” Still no turtles. We sat anyway. He got out his binoculars and scanned the trees across the pond. Then he gasped. “Oh, my goodness… there’s a bush in your way, come over here and look.”

So I picked up my binoculars, scooted over, and looked. On a sandbar near the opposite shore sat over two dozen turtles of various sizes, sunning themselves in a decorous row. One does not normally think of turtles as social, but that was quite a shelly conclave.

From Turtle Pond, which had nobly lived up to its name, we walked up Monroe Street and from there to Vilas Park. There we detoured to go see some Canadian goose families. Canadian goslings are as pinfeatherily cute as all young waterfowl. Their distinctive feature is their golden-haloed heads; all the feathers that will turn black in the adult goose shine dark gold.

Though the zoo was crowded—I did say this was the best weekend day in weeks—we enjoyed ourselves. The lemurs and the Colobus monkeys carried on their cold war (over proper tail configurations, no doubt; I don’t believe they will ever settle it). The year-old orang, still a baby, swung on things and caught a piggyback ride on its mama. One of the peacocks decided to put on a show for everyone—only the second time I have seen one of them do that, and we’ve been going to the zoo for years. The misnamed Herpetarium had its usual collection of nifty fish, frogs, snakes, bats, lizards, bugs, and miscellanea. (I am fond of stick-bugs, myself. Times I wish I had camouflage that good.)

We also said goodbye to two tigers, born here a few years back, who are moving out this week to a zoo in New York. Unfortunately, their bloodlines are not pure Siberian, as was thought when the decision was made to breed their mother. Still, they were wonderful babies, grew into lovely beasts, and will doubtless capture as many hearts in New York as here.

Our zoo tradition is to spend the day there, then have dinner at Pascual’s and dessert at Michael’s Frozen Custard. (Custard is the real food of the gods, by the way. Ye who know not custard know not heaven.) By the time we got there, though, I was limping; I stupidly decided to try out some new walking shoes, and they don’t fit worth a damn. We walked up to Regent Street to catch a bus home.

Today I am a touch sunburned about the forehead and nose, badly blistered about the heels and toes, and the odd way I ended up walking left bad memories in my calf muscles—but the peaceful, pleasant day I spent was worth all that and more.

My Kinesis Contoured keyboard spent the week at work, and then came home with me for the weekend. I am so attached to this thing that I turned the closet inside out this morning looking for a waterproof-able bag large enough to fit it into, to make carting it back and forth from work easy and safe. (Safe for the keyboard, I mean, of course.)

It hasn’t cured years of damage in a single week. How could it? But it has already helped, and I think it will do the trick, long-term.

Which is not to say I like everything about it. The function keys are too small, and I would really dig a set of normally-configured arrow keys in the center somewhere. (The Maltron keyboards seem a little saner in this respect, but they are more expensive and decidedly larger and heavier to boot.)

Since it’s a given that any user of this keyboard is going to remap some keys, a duplicate set of key labels would be a real help. Not all keycaps can be exchanged. I remapped the Ctrl and Shift keys on the left-hand side to spare my much-abused left pinkie. The keys, however, are now mislabeled, because the Shift keycap doesn’t fit in the Ctrl spot. I have gotten sufficiently accustomed to this layout that in normal typing I’m fine, but coding (HTML, CSS, or Python) is still difficult.

Do not buy this keyboard if you can’t use it for at least 90% of your normal typing load. If you look at a normal keyboard layout, the keys for any given finger are laid out diagonally. This is a remarkably stupid arrangement, since our fingers aren’t designed for lateral motion. The Kinesis Contoured straightens out the lines, which is lovely once you get used to it. I am touch-typing numbers accurately for the first time in my life. Once you are used to it, however, touch-typing on a regular keyboard requires considerable re-acclimation. After a while, I am sure I will hardly be able to do it at all.

Fine by me. I am tired of pain. If carting this keyboard around gives me my hands back (who knows? perhaps someday I’ll be able to play my tenor recorder again), I will happily cart.

The macro-programming ability is super-keen, and refreshingly easy to use. I am using it at work to keep from retyping yard-long place descriptions that the Puerto Rican census enumerators used to locate themselves in rural areas. At home, it is handling awkward markup-related typing chores (such as the &# combination that starts Unicode character entities, which I use to get typographic quotes and en and em dashes because I am just that much of a typesetting geek). All macros are stored in the keyboard, so they survive being moved from computer to computer.

I am less thrilled with the footswitch thus far. I suspect that people who drive regularly, and are used to creating a light, constant touch on foot pedals, will have an easier time with it than I have had. The workable configuration for me turned out to be turning it around backwards and hitting the pedals with heel rather than toe, and even that gets wearisome after a while. Programming the switch is easy, at least, and like the keyboard, the switch can be moved between computers without losing previously-programmed actions.

However, the programming software provided is the likeliest culprit in a very weird problem I had on the Win98 machine at work: the left mouse key suddenly quit working. Mouse still moved, right-click still worked, but left-click went completely dead. Reboot machine and everything was fine for a while; then left-click died again. Repeat until disgusted, then remove footswitch programming software; all is then well. Fortunately, the programming software is only necessary to actually program the footswitch. Once programmed, the switch works fine without the software.

I don’t mind saying that I was scared stiff of not being able to continue in my job because of the pain it was causing me. I am not afraid now. Gonna be fine, in a while.

Jenny, as usual, scoops me on ebook stuff: this time, a sober examination of the current situation and prospects for ebooks.

Also on Jenny’s recommendation, I picked up Crossing the Chasm from the library. The combination of these two readings produced some curious ideas about what’s gone wrong and how to fix it (if it’s not already too late, and it may be). These are very rough ideas still, but there might be the germ of something good in them.

Crossing the Chasm trumpets repeatedly the need for new technologies to penetrate highly specific markets before making the move to the masses. This focuses development effort and provides a word-of-mouth base for future market expansion.

This was quite feasible for dedicated readers. Quite feasible. In fact, the dedicated-reader success stories I heard followed precisely this pattern. Airplane reading material. Quick dissemination of newspaper-route information. Manuals. None of that—not even the airplane material, though it comes closest—is mass-market.

What did the dedicated-reader and content-library boys do with these successes? Ignore them, in favor of wooing the Big Publishers with their multi-million sales figures.

Mistake. Big one. Suddenly the ebookers were wooing two markets at once: publishers and mass-market readers. They did a lousy job at both. Publishers needed help with production workflows, and wanted assurances against hacking. Readers needed immense numbers and varieties of books available before they would take the plunge. Publishers needed production standards. Readers needed end-user standards.

The ebookers responded to the production challenge by trying to take over production. They weren’t equipped for it. They didn’t know squat about book production. They didn’t know squat about markup. They just thought they could make money either by using production as a loss leader for exclusive content acquisition, or by retailing the end-product, or by retailing the platform on which the end-product was to be read.

Didn’t none of that work, because book production is not amenable to mass-production practices, and good book production is harder still. What the ebookers ended up with was hugely expensive and unwieldy production systems that produced lousy results, and publishers who eventually wised up enough to be furious at having their content hijacked merely for the price of production.

Meanwhile, the readers wondered where all the content was, and why it was so dang hard to get to. Where was a standard ebook format? Ebookers, a few of them, pointed to the Open eBook Publication Structure. That, however, is a PRODUCTION STANDARD, absolutely NOT AN END-USER STANDARD.

The ebookers ended up in the weird position of showing a production standard to publishers who had utterly abdicated production processes to third parties. Publishers allied themselves with readers in asking for end-user standards, because they understood they needed to reach as many readers as possible. Since they didn’t understand production, they didn’t understand that end-user standards are a big fat RED HERRING, only somewhat less likely to exist in nature.

(Why? In a word, obsolescence. Production standards, if well-crafted, don’t have to become obsolete, and so they enable content created by them to avoid obsolescence as well. Whereas there is no such thing as a non-obsolete-able end-user standard.)

So nobody was happy.

What should have happened instead? Well, if the whole problem came about through trying to please two widely variant groups—producers and consumers, publishers and mass-market readers—how about finding a situation in which producer and consumer have tighter ties? How about the ideal situation, in which (in the aggregate) producers and consumers are the same people?

Do such situations exist? Sure they do. It’s no great wonder that science fiction has done comparatively well in ebook form. The explanation is not technogeekitude; it’s proximity. Science fiction readers and science fiction writers are a lot closer to each other than are readers and writers of typical mass-market or even midlist fiction. They meet at cons and shows, write to each other, hang out on the net together. Science fiction writers also read science fiction, lots of it. I defy you to find a science fiction writer who didn’t start out an avid reader.

The real target, though, the one the brainless mass-market-fixated idiots at Gemstar and Microsoft and Cytale and Mobi and Palm and Adobe and bloody well everywhere else missed, is academia. Is there a more incestuous readership anywhere on earth? Even more invitingly: is there a book production system anywhere less concerned with immediate profit, and more concerned with easy dissemination? Is there a system less needful of complicated DRM? Is there a readership anywhere with more immediate prestige? Is there a readership anywhere less enamored of the print production process? Are there readers (save of course librarians) with more influence over the reading habits of others?

This should have been a no-brainer. Grab the professors, and the world will follow. But nobody thought of it, because nobody saw immediate dollar signs. So they went after a market that when all is said and done is perfectly happy with paper, and they got their clocks cleaned for it.

What really makes me angry is that they took the immense potential of electronic books into the gutter with them. Makes me furious.

So I owe Frank Paynter an apology. He emailed me earlier this week to say that he’d run into my blog, his office is just down the street from my home, isn’t that freaky, and would it be okay if he called me? Sure, I said, here’s the number—and then I let my husband dial in from the home line all evening so Frank couldn’t get through. And then in all the anniversary fuss, I never told Frank what had happened.

Totally my fault, Frank. I apologize. Try back? I have him connected to the dialup line now. Promise.

Today I got an email from Tom Shugart, who has a kid at the UW, attended my alma mater (Hoosiers rule, Boilermakers drool), and writes a lovely blog. He sums up the experience I had this week nicely:

One of the most rewarding parts of blogging is when you discover a great blog–not through surfing–but through the receipt of an email from the blog’s author telling you that s/he enjoys reading your blog. You have no idea how this person came into your orbit, but there s/he is. It’s a wonder.

This is yet another “how I’m going to do it” post, but with luck there will be a juicy bit or two.

We now have three blogging tools in play, so it seems pointless to work only toward a Blogger template. In any case, checking our results in a browser if all we have is a template is awkward to impossible. Here’s why: the blogtool-specific tags are only occasionally XML/HTML compliant. Any blogtool tag you see with a $ or two in it is not XML; it just looks kinda like it.

This means that it’s impossible to validate a blog template with a typical XML/XHTML validator. You can only validate the results of the template, that is, what happens when posts are fed into the template by the blog tool.

It didn’t have to be this way. XML contains a gizmo known as the “processing instruction” specifically designed for this sort of situation. None of the blog tools I know of use it. Go figure. (What does a processing instruction look like? Well, the <?xml declaration we included and then excluded from our markover template is one. Processing instructions start with <? and continue with the name of the processor at which the instruction is aimed plus a space. They can then contain anything you please except a >, which is the closing character.)

So what I’m going to do is use some of AKMA’s old posts to set up a page that looks like what the final blog will look like. When we’re happy with that, I will translate it into a Blogger template, a Movable Type template, and (assuming I can find sufficient documentation) a Radio template. That way, whatever tool AKMA chooses, he’s covered.

Since I run MT myself, I will also set up a separate AKMA test blog (and give AKMA access to it), to be sure I am handling the template correctly. If there’s a Radio user willing to do the same, I would surely appreciate it.

Mark Pilgrim is demonstrating the uses of HTML tables for a table art contest.

Nice Mondrian he’s got there. Could one try for a fractal table? No, I know—there’s this local guy name of Marko Spalatin (hope I haven’t mangled his name) who does stuff that would be perfect…

Oh, and happy half-birthday, Mark!

(Update: Mark, you are one sick puppy, that’s all there is to it. View Mark’s fractal table with appropriate awe.)