MarkBernstein.org
Dec 02 4 2002

Cutting Problems

We've been discussing information architecture (IA) and the state of the Web. Are most sites lousy, as the IA textbooks say? Or are most sites pretty good? Let's take an anecdotal but, I think, typical look at how good commercial Web sites are these days -- even little, out-of-the-way merchants.

Last night, I was rereading part of Kitchen Confidential, in which Anthony Bourdain was talking about the importance of chef knives and the new preference, among professionals, for lightweight Japanese knives. How much did these cost? I tried asking at my local kitchen store last year, but they were clueless.

So I popped up Google, entered the brand name, and went to the first retailer on the list, a place called ChefKnivesToGo. I found a knife that sounded like the one Bourdain's sous chef likes; Bourdain's favorite was a little too expensive for a midnight impulse buy. I bought it. There's email in my inbox this morning; the knife is on its way.

Now, this site isn't remarkably beautiful or efficient, but it's pretty good. In fact, it even offered help show how I ought to use that new Japanese whetstone my college roommate got me to buy this summer, which is nice because the stone comes with instructions in Japanese. There are some visible mistakes, some dubious design decisions, and tying your brand so prominently to cerulean blue might be a mistake. But none of this matters much -- not compared to having the knife Bourdain's sous chef likes.

The proprietors of this site, it turns out, are Mark and Susan Richmond, a Wisconsin couple who used to run a futon store in Madison. After they cashed out of the store, Sue went to culinary school. They started selling knives last August. They keep stock in their home.

Could I design a better site? Probably. I've been thinking about this since 1982, they've been thinking since August. But I'm not certain my site would make more money. Are you?

Was the site usable? Probably. We'll know in a couple of days, if the knife arrives. Sure, they could have saved me a minute or two. Give or take a stoplight or an elevator wait, it's no big deal. (Might be a big deal to a big store, amortized over a million transactions, but we're talking about user experience here, not retail efficiency)

This is unscientific: I might have been lucky and run into an unusually good site. But I don't think that's the explanation....

Information Architect Christina Wodtke (author of the excellent introduction, Information Architecture: Blueprints for the Web) responds to my probe, "What's your Problem?"

Mark Bernstein thinks we are saying the web sucks. Er, well....

.....

But there one heck of a lot of bad websites out there..... the grand mass of *professional* sites are lousy.

I'm not even including the sites built by hobbyists, families and diarists. These are small busines sites whose equivalent is a store in the mall, or on main street in a small town-- but they don't even meet that standard. You can't find anything on these sites, they are ugly as nobody's business and check out is often impossible to accomplish.

What standard should a Web site meet to escape the opprobrium of being deemed

... like a old barn turned junk store you find on a lonely country road... you dig through spiderweb covered junk, and if you do find something that catches your eye, you pay in cash because the cross-eyed KKK-T-shirt wearing drooling kook at the register.

This isn't saying, "There are some things here that, I would argue, are not optimal." This isn't saying, "There are some things here that are demonstrably wrong -- things that could clearly be clearer or more convenient. (My grocery sometimes uses quotation marks for emphasis: this is illiterate, but I can manage to find the fruit anyway.)

I read a lot of Web sites, and a buy stuff from all sorts -- including some very small sites. Unless I go looking for those cross-eyed kooks, I don't seem to run into them all that often. Curious.

I've been reading a lot of Information Architecture lately, and one idea is weirdly pervasive -- the notion that most Web sites are bad. Everywhere you look in the literature, you see warnings about unusable sites, idiotic sites, disorganized and chaotic sites. Sites that suck.

Is this true? Does anyone really believe this?

Yes, you can find bad experiences on the Web. You can find them right outside your door, too. Life is like that. Don't blame the Web.

Yes, it's easy for any Web pro to find things that could be improved. That's always true, in anything. Show a rocket to a rocket scientist, and she'll tell you how to make it better. So?

Information Architects sound as if the Web is terrible. That's not my experience. I bet it's not yours, either. Perhaps it used to be bad, back in the day. (I was there, and it wasn't that bad, but never mind) It's not that bad now. In fact, it's great!

Why do people say this? Perhaps, a few years ago, it was more tactful to tell your client, "Your web site is poorly designed" instead of the blunt truth, "You have no business model."

Could the Web be better? Of course! Could our sites be more convenient, faster, smoother, more profitable? Naturally. There's always more work to do.

Trying to establish a profession on the foundation of a myth is, I think, a tactical error.

Enjoy what we've all built together, dream of what we can do next. Have a happy Thanksgiving. Enjoy! See you next week.

The new Digital Storytelling Association is an outgrowth of the old Digital Storytelling Festival, which will be revived this year. Their fresh new website includes a newsletter with my own survey of The Year In Literary Hypertext, a story by Brenda Laurel, and more.

Nov 02 22 2002

Bumper Sticker

"Your choice: Support the Democrats in 2004, or support the Resistance in 2006"

Nov 02 21 2002

Time

If you're interested in narrative time in new media, run -- don't walk -- to today's Bruno. (The main page is here; the permalink for this week is here)

I don't think Christopher Baldwin spends a lot of time with Genette in any season, and he's not in a theoretic mood right now, since he "just got some news yesterday about something....and it sent me spinning again." Even so, take a look at this page and start counting the number of different kinds of time on the page.

What does it cost to run a news site or a big weblog? Robin Huff, editor of Mac Net Journal, estimates $600/month, for hosting and labor.

This is probably a low estimate, because the labor costs seem artificially low and because he omits marketing, equipment, and overhead. Huff's other business can subsidize the Journal, but it's probably a good idea to consider the value of this subsidy. Huff can use the same desk for writing the Journal and for his main business. If he weren't using the desk for the Journal, though, it might just sit idle: it's hard to rent desk space for a couple of hours a day. So it doesn't cost him more to provide office space for the Journal, but the office space still has value. It's like buying eggs: you've got to pay for a whole egg, even if you only need half.

New technology can change this. For example, consider the business role of message technologies like SMS and iChat. Normally, workers are like eggs: if you want a sysadmin, it's cheaper to buy a whole sysadmin than a fraction. But this is changing: I think it's entirely practical for several small companies to share one sysadmin, using cell phones and the net to make a single sysadmin nearly as available when needed as separate staffers would be.

Torill provides lots of useful links to discussion of narrative in games.

There's some good work here, but also a lot of academic game-playing. Either games can hold nature's mirror, or they can't. Either games can tell us about love and sex and mystery, or they can't. Either we want to meet Hamlet on the holodeck, or we don't.

And until we have real criticism of real games, nobody will know. Where is the good game criticism?

Nov 02 19 2002

Linkage

ALA editor Jeff Zeldman observes that, "According to its ALA popularity index, Mark Bernstein’s “10 Tips on Writing the Living Web” is the most-linked feature A List Apart has ever run. Wow!

Nov 02 17 2002

Reading Tools

Doug Miller discusses his Library Of The Future, tying together Tinderbox, Google, Amazon, and e-books. It's not a Vision Of The Future, but a hard-nosed approach to managing your time and ideas right now.

"The combination of Amazon, e-books, and a personal content management system like Tinderbox is allowing me to create what I call my "Library of the Future." I'm using Tinderbox to set up a sort of master "card catalog" of the books I'm reading, have read, or intend on reading. Each book has it's own note, containing appropriate meta-data ....By using Tinderbox as an application to "glue" different, knowledge-rich information sources together, I've created a system that allows me to easily discover, purchase, track, annotate, relate, and search my important books in a "just-in-time" fashion.

Thoughts on seeing a fine Harvard student production of Cabaret :

  • Choosing to do Cabaret at this particular moment is a political statement. It might turn out to be a brave thing to have done. We'll see.
  • I'd not seen Cabaret on stage before. It works. I've never read any Isherwood, either. To me, he's an engimatic character who first appears in Cabaret, and then is glimpsed much later in Tales Of The City. I ordered a copy of Berlin Stories from Amazon; it's raining, WordsWorth might not have it, and where will I find time to read it, much less schlepp back to Cambridge? (The book world benefits by getting books to people. End of discussion.)
  • When I was a TA, I thought Harvard undergrads spent way to much time on extra-currics. This is the other side of the coin: the production is knee deep in talent. The band was superb all night. Alex Schemmer as the Master of Ceremonies saunters on stage, cocks his hips, and suddenly you aren't thinking of Joel Grey anymore. Meredith Schweig (Fraulein Schneider) seems to have wonderful voice control, acting and singing alike; she doesn't know what to do with her body, but that will come. Adele Jerista makes a fine (though short) Sally Bowles. Linda and disagreed about her rendition of the title song (which appears once in this production, late in act II, standing before a mic on an empty stage). Linda thought it tried to hard not too be Liza, but I thought it worked just fine -- and Jerista looked like she was having plenty of fun, almost 3 hours into the performance, and would be up for drinks afterward, ok? Let's play two!
  • We got 3rd row seats because the Loeb is notoriously tough for student actors (and, in fact, the American Repertory Theater came to Harvard because Harvard built the stage and then realized that it swallowed amateurs). Needn't have worried. First, the production was mic'ed to the hilt. Second, they're all belters, anyway. Come HEAR the music play!
  • I grew up on the North Side of Chicago in the 60's. Second City was a a few blocks from home, Organic Theater and Kingston Mines (where Grease started) were around the corner, Steppenwolf was starting nearby. I figured, neighborhood theater was normal. I miss it.
  • A central problem in the performing arts of our time remains Joni Mitchell's fellow who played real good for free. Because we can flip a switch and see the very best, how do we enjoy the good? The flawed but brilliant? The erratic artist having a really good night, or the vet who has lost a step but can still deliver? They knew he'd never been on their TV, so the passed his music by. One hint that we're learning how to do this might be the revival of minor league baseball. We'll see.
Nov 02 16 2002

Experienced

Jesse James Garrett's new Elements of User Experience is pleasant. Nobody will be offended. Nobody will disagree very strongly. Oh well. In Books.

I bumped into Michael Joyce and Carolyn Guyer at Hammersley's (g) last night. Linda and I were having a rituaL feast. Hammersley's makes a truly excellent roast chicken. A signed and numbered limited edition of Joyce's new hardcover, Liam's Going, is available at Eastgate.

Diane Greco's new blog design is hip, fly, and gray. Her baby, though, is getting an early start on losing things.

Torill shows good business sense when she reminds us that this whole business about bookstores and fast food is a red herring. The book world is not the province of the rich; there's a complex class issue bubbling near the surface.

Fast food is driven by fear of the unknown: if you're someplace new and you have to find something to eat, a chain restaurant is safe. It won;t be weird, it won't be unsanitary, you won't be shocked at the prices or the ingredients. This has nothing to do with books; books seldom give you gastritis.

Bookstores (and libraries) succeed when they have what you want. They have two strategies for doing this: selection and size. Either they cleverly guess what you want, or they have a big selection. That's the core of retail bookselling.

You can't guess right for everybody. If you are in New York or London, you can focus on a core group and guess right for them. Stuart Brent built a successful bookstore in the 1950's by focusing on Chicago-area psychoanalysts; his book about the experience, The Seven Stairs, is well worth reading. You'll find this strategy underlies many legendary stores. But this can only work in big cities, because you need population.

If you can't guess right, you can stock a lot of books. The superstores attempt this. It's a tough strategy, because you need to pay rent for the space those books occupy, and you need to pay interest on the books you haven't sold yet.

If you want a book world that is based on small retail stores, the only books they'll be able to sell are the handful of books that are sure to sell lots of copies in Des Moines and Dallas. That means a small selection of the latest quack medicine, quack religion, celebrity gossip, and formula fiction. Bookstores that are worse than airport bookstores -- because airport bookstores have, by comparison, an elite, educated, and wealthy clientele.

Nobody addressed my key point, incidentally, which is that internet orders can't help independent stores much, anyway. They've got less than a 40% markup to play with in the first place, and most independents are barely squeaking through on that. Now, in addition to all our current expenses, we've got a larger transaction cost, a share of the Web infrastructure, the time and materials for picking and packing, fees for special orders, fees for support, allowances for fraud and lost shipments and returns -- and that's in the best case, where the book is in the store, gathering dust on the shelf! If we need to go to Baker and Taylor for an extra copy, we've got two more transactions to reconcile. We've got to be ready to field calls from customers, answer their email, respond to their letters. Amazon can manage (maybe) through scale, through sidelines, through very specialized wholesaling and warehousing systems. They rent warehouse space in rural Washington and Delaware, not prime commercial retail in Manhattan.

Perhaps I'm missing something here. I'm not a bookseller. But I think we're talking about a pipe dream. (That said, it should be really easy to expand any well-designed aggregator to interpret booksense links as well as amazon links)

Nov 02 14 2002

Lifeblood

Jill has a fascinating, rambling discussion of bookstores, dialogue, and weblog form, responding to Noah Wardrip-Fruin's concern that Amazon links help clobber independent bookstores, "which have been a lifeblood of our culture."

Amazon links may at least help Amazon prevent Barnes & Noble (or Time-Warner, or Bertelsmann) from becoming a monopoly.

Retail book distribution is broken. It's been broken for a long time: the trade practices of the retail book business were frozen in the early years of the Great Depression. In much of the US, there are no really good bookstores within reach. In much of the US, there have not been any really good bookstores for a long time.

Noah likes Booksense. Running the numbers, I have trouble convincing myself that Booksense can make a real difference to the survival of any independent store. There's just not enough margin in the system. The case is marginal for Amazon, with its economies of scale; it's got to be worse for Booksense. You've got short discounts (40%, often less), unpredictable demand, lots of small, expensive transactions, and abundant competition. Retail bookselling is a tough racket.

Steve Johnson (Emergence) has a shiny new weblog. Jill/txt is on his blogroll.

Nov 02 12 2002

Characters

Jill Walker offers some thoughtful speculation on hypertext with characters. She prizes dialog for its "questions, uncertainty, different points of view and an evolving sense of the answer".

Do any of the new spate of "information architecture" books address this?

Eric Scheid responds to say that:

"I've been reading a spate of IA books lately, and so far I haven't seen any indication at all. They are all pretty much written with the base assumption that information retrieval is a matter of destination, not journey. That is, the goal of IA is to facilitate the fastest, clearest, most direct, least distracting route to the information wanted. A corollary assumption is that people know exactly what they want to find.

Confidential to the IA community: people will tell you, if you ask, that they know exactly what they want. If you believe this -- and most protocol studies seem to -- I have a nice bridge in Brooklyn for you.

The Pentagon is writing programs to read your email and correlate "suspicious" activity with telephone, banking and credit card transactions. Of the official in charge of this program, the NY Times writes:

"Before taking the position at the Pentagon, Admiral [John] Poindexter, who was convicted in 1990 for his role in the Iran-contra affair, had worked as a contractor on one of the projects he now controls. Admiral Poindexter's conviction was reversed in 1991 by a federal appeals court because he had been granted immunity for his testimony before Congress about the case."

"Give the right-wing Republicans credit. What they stand for is wrong and dangerous, but at least they know what they believe." -- Dan Gillmoor.

Hypertext research pioneer Norm Meyrowitz bylines the vision statement for Macromedia's new Web editing tool, Contribute. In some ways, Contribute recalls Norm's HT89 keynote proposal on Warm Links, a topic nicely summarized here by Michael Bieber et al.

Nov 02 10 2002

Lost Lit

The slow drift of time and language means that, sooner or later, all art is likely to need footnotes. The process is strangely swift, though, for a number of relatively recent works that suddenly seem antique -- and almost incomprehensible -- because they speak in codes we've forgotten.

Take Suddenly Last Summer, It's a good movie, a fairly famous one. Katherine Hepburn and Elizabeth Taylor both perform well. A play by Tennessee Williams, adapted by a young Gore Vidal, directed by Joseph Mankiewicz. It's about a victim of gay bashing, but you couldn't say that in 1959. Instead, you used all sorts of codes, and the codes are getting hard to decipher. Same thing with The Naked And The Dead, the famous first novel of a definitively famous writer. The whole plot is pegged to Gays In The Military, right? Does anyone ever mention this? I mean, we're twenty years before Stonewall here, and almost fifty before DontAskDontTell.

Nov 02 9 2002

MNJ in Tinderbox

Mac Net Journal has switched to Tinderbox. Welcome.

In Card Shark and Thespis (pdf) (flash), I argued that, in a first-person interactive story, we must not be the hero-protagonist, that we must instead be weak and unimportant.

Illusions that place the reader on stage necessarily founder when promised freedom of action is contradicted by the limitations of the simulated environment. IF asks us to find a creative, imaginative, and successful resolution to the dramatic problem. The imaginative reader is bound to think of things the creator never envisioned, and the reader's best thinking inevitably generates the dullest response: "I don't understand." The computational environment can never match our aspirations, and allusions to unlimited computing power of the future (the starship holodeck) can't rectify the fundamental problem: readers will always want to do things nobody (and no computer) could anticipate.

Another World seems to contradict this: in Another World, you're the powerful Resident, the chief representative of the all-powerful Company. You say, "Build a market here!", and a market is built. You say, "Expand the archaeological work, and put a hotel and museum by the dig", and tour companies spring up overnight.

But your budget and your authority don't supply easy answers when two love-struck kids (an elf and a human, for pity's sake!) intrude on a demon Initiation Rite, or when aboriginal ideologues threaten a general strike, or when you just can't get that irritating reporter out of your office. You've got power, you can do lots of things, but its the wrong kind of power.

(Another goal of Another World is to show that the Unimportant Player conjecture -- the argument I call "My Friend, Hamlet" -- applies beyond Tragedy. )

(more on AnotherWorld)

Stanford Magazine reviews Richard Holeton's hilarious hypertext, Figurski at Findhorn On Acid.

"Readers of conventional fiction, beware: there are 147 scenarios in Figurski’s brainchild, and not a single one can definitively be characterized as the beginning or end. But Holeton, who has always admired the works of genre-elusive authors such as Vladimir Nabokov, James Joyce and Italo Calvino, is out to create a different read—an experience as fragmentary and innovative as the Internet itself. And he seems to be succeeding."

Eastgate is looking for someone who can lend an occasional hand with Web projects. Light Perl, PHP, CSS, and simple BSD server configuration -- a few hours a week. You can be anywhere in the world, though it helps if you're well wired. Tinderbox a plus. Ideal for a Web independent or small firm who want to make decisions endless meetings and to get involved with visible projects that make a difference. Email bernstein@eastgate.com, or iChat EastgateSystems.

Nov 02 6 2002

Solid

I froze the system of the Tinderbook this morning, and had to force a restart.

What's notable is that I hadn't restarted the Tinderbook in ages for anything besides software installation, configuration, and the like. We're talking several weeks here.

That's remarkable. I use new, untried software. I develop software. I'm constantly adding new stuff, I crash applications many times every day. I'm always carting the computer across town.

Another World has to be put together quickly, using widely-available parts. It's not going to be a small project, but the budget for time and tools needs to be small.

The main engine is built in Macromedia Flash; I'll talk more about this later. The data files -- including all the writing, is stored in a set of XML files. (Another set of XML files holds user information and games in progress) The XML files, in turn, are created with Tinderbox. There's no special support for this in Tinderbox; I just whipped together some simple XML templates and Bob's your second cousin once removed.

The nice thing about this is that I get a decent writing environment (Tinderbox) that creates nicely parseable files that Flash can read. Maintenance is easy, and we don't need to worry about arcane specs or building a game editor. (more on AnotherWorld)

"Well, that really could have gone better. " -- Joshua Micah Marshall, Talking Points Memo ( a remarkable weblog on the US political scene)

Nov 02 5 2002

Hyperfriction

"Hyperfriction occurs when the hypertext opposes the reader....Hyperfriction arouses desire. To enter the text. To know it."-- Anja Rau, Flickwerk

It's election day, and once again I'm hoping that the fairly-good candidate will beat the horrible candidate. It'll be a close race.

The last time an American liberal or progressive could cast a vote for a presidential candidate whom they could wholly support and who had a chance to be elected was 1960.

Nov 02 3 2002

Microsoft

Gillmor despairs for the future of software innovation, now that anti-trust law is seen to be useless.

Since technology will be a piece of everything we touch, get ready to find Microsoft's unclean fingers in every pocket and wallet.

The one chance is for people to realize what's at stake and do something difficult: Make choices that mean less convenience today in order to have liberty tomorrow. Americans are lousy at this, but a lot is now at stake. You may not care. You should.

Nov 02 1 2002

User Experience

Jesse James Garrett's intriguing new book, The Elements of User Experience, repeats a common slip you see all the time in the design literature. He begins with a horrible chain of events -- alarm clock malfunction, coffee maker fails to make coffee, gas station has long lines and slow service -- and traces all these annoyances to design problems. Then he writes that:

"Every one of the previous incidents of 'bad luck' could have been avoided if someone had taken more care in designing a product. These examples all demonstrate a lack of attention to the user experience.

Attributing the problem to carelessness and inattention is common in design books. I think it's wrong.

First, great design is only easy in retrospect. It's like the chess problems you see in newspapers: once you see the answer, it's obvious, but until then the problem is often a real challenge.

Second, it's probably wrong to assume that mistakes arise only because people weren't paying attention, or didn't care, or were conceited, self-indulgent idiots. Mistakes arise because people were avoiding other mistakes, because they were trying to save your time and money, because they're human.

Jesse James Garrett has a cat that likes to walk on his the alarm clock at night. The designer of his clock didn't anticipate this. Is the designer stupid, or merely someone whose cat sleeps downstairs?

You see the same assumption in the business press all the time. Winners are brilliant, insightful, witty. People who didn't win are dolts and fools. This is CEO porn. It's sometimes true, I guess, but life's not like that.

Same thing with sports, by the way: people act like professional athletes are little leaguers, that everything can be fixed if they simply want to win very badly. Sure, motivation matters sometimes, but having the bigger battalions matters all the time. It's not a matter of Barry Bonds wanting to win the series; having a bench matters, too.

Confidential to editors everywhere: business people know what return on investment (ROI) is, and why it's important. If they don't, you can't help them. You don't need to spend pages 13-18 explaining this. Really. Even if the author wants to.

Rei Shimura, young Japanese antiques dealer, returns to the country of her birth as a museum courier. Sexy, unsophisticated, fun. in Books...

How can Another World be exciting? Since it's a hypertext, we don't know how people will begin, or where they'll go. There are few borders. But there should be plenty of tension.

There's the interface -- a new mechanism to explore. New games are always exciting to people who play. (Some people despise anything that looks like a computer game, but they won't like Another World anyway).

There are the characters. Jules, the beautiful, shy young elf. Roma, a farmgirl who would love to take him to bed. Roma's father knows his Shakespeare, but doesn't want his only surviving daughter to make a terrible mistake. What will they do? Can we watch?

Outside town, there's an archaeological dig in progress. Rumors say, they've found something important. Rumors say, they'd better work fast, because the Administrator wants to build a new development on the site. Rumors say that one of the grad students working on the site is having an affair with the leader of the Resistance, behind her Professor's back. It's a small town. There are always rumors.

Jules' friends want to talk to him about doing some favors for the Resistance. And then there's Miles Corbett, correspondent for the Times and rumored intelligence agent, who has a camera and who has been spending a lot of time near the glade where Jules and Roma have been sneaking stolen moments.

(Of course, in some readings none of this might happen. If you decide The Company needs to build a new mine right where Jules goes bathing in the twilight, then Roma won't ever come across a beautiful young elf, naked in the stream.) (more on Another World)