It's all spinning wheels and self-doubt until the first pot of coffee.

Bitterness from the halls of Xanadu

I'm a little late on this, but I just read an BBC interview with Ted Nelson this weekend. I don't know the man, and am familiar with his work in only the most cursory way - I've read a bit about Xanadu, skimmed some articles on its history, but I've yet to download the code finally released a few years ago and see what's what. Having said that, this interview reads like the bitter mutterings of a guy who wants to slug TimBL and the rest of us web hackers for making him all but obsolete.

From what I've read, the body of work surrounding Xanadu seems to have a lot to say for itself, though some of that - the assorted collection of almost psychedelic jargon invented to describe its convolutions - seems almost self parodying. The history of Xanadu and Nelson's work with hypertext systems looks to me to be yet another proof that ideas are fucking worthless, and another vote in favor of bazaar- over cathedral-style development.

Maybe it's the fault of the interviewer, but Nelson comes across as a bit self-aggrandizing in trying to puff up how creative and multi-talented and immersed in media - I mean, he made his own rock opera for cripes sake! So I guess this should set the stage for his authority when he dismisses HTML as "a degenerate form of [hypertext] that has been standardised by people who, I think, do not understand the real problems" and the web as "... trivially simple ... massively successful like karaoke - anybody can do it."

But y'know, for all his criticism of systems here and now, and claiming that the people involved with the web don't know what they're doing - he has a surprising lack of working software out in the world. That's what I consider success - working and well used implementation. Who doesn't understand the real problems? One can only build architectures of vapor for so long up in the ivory tower before one must pit it against the world. I don't suppose he's ever tossed around the idea that maybe, just maybe, Xanadu hasn't stormed the world because it's too big and cumbersome and amazingly convoluted for anyone who hasn't worked with it for 30 years to put into use? There is much to be said for the benefits of Karaoke-simple technology.

So yeah, maybe on some Platonic plane of ideal forms, compared to The One True Hypertext, HTML is crap. Okay, maybe right here on my hard drive, HTML is crap. But it's on my hard drive, and I use it. It's been learned from, and attempts are being made to improve upon it. As far as I know, Nelson's ideas of perfection have never seen a pounding from the world of imperfection. That's the crucible in which things are really formed. You get your software in reasonable shape, toss it to the wolves, and see how it fares. Realize that some of your prized pet theories and designs were bullshit, and rework the thing for the next time around.

Okay, that's my rant. Now I'm back to work, and possibly maybe off to check out more things Xanadu and Nelson on my lunch break.


Archived Comments

  • For what it's worth, Nelson inspired a generation of software developers with his book Dream Machines, which came out in the mid-70's. He deserves a lot of credit for that.
  • There is, of course, that - the "Ideas are fucking worthless" meme needs some qualification. They do have worth, just not so much worth as execution of ideas as far as I'm concerned. From what I've read (not having been there in the mid-70's), Nelson's been influential like crazy. He'd envisioned so many of the things taken for granted today - but every time I read an interview with him, he seems so bitter with all the imperfect renditions of his ideas. Seems like he just can't get around to accepting that "pretty good" implementations are good enough. But, I could have him all wrong, having gotten all my impressions from mostly second- and third-hand sources. Thanks for the book reference, though - must read that someday, although a cursory search seems to show that it's a rare find.
  • While there is bitterness there, I read it more as long term frustration. It looks similar to the attitudes I think I've seen in Jef Raskin and Doug Engelbart. These are people with a clear personal vision of some aspect of computing that they feel is radically better that the existing (flawed) practice. This happens all of the time, you say? No big deal? Sure, but in their cases there are factors that make it much more frustrating. In no particular order: Their vision predates and was the _misunderstood_ inspiration for the existing models. Sometimes it predates them by a lot. Can you imagine thinking of this stuff in 1962 then having to wait 25 years for people to start getting it? Or to give a demonstration in 1968 of a working intelligence augmentation system (http://sloan.stanford.edu/MouseSite/1968Demo.html)...and then being best known for the mouse. Or to be best known for creating the Macintosh, which was then ripped from you while your better later work is marginalized. Doug and Jef _have_ produced working systems. They've usually been misunderstood, or ignored. One of the hallmarks I've noticed in all their works is that they are systems thinkers, whose vision works best if it's taken altogether, not piecemeal. TBL's Semantic Web fits into this category as well. There are synergistic effects between the components that should make their total effect much more than the linear sum. I suspect that this is why they tend to see 'pretty good' versions as not good enough. There is a threshold level that has to be exceeded before it is truly good enough. They all seem to see their visions as fundamentally simple and often it is a source of great confusion as to how others can continue to not understand. In the past they haven't received much credit for their work outside a narrow group of insiders. This has changed to some extent but it's hard to become all happy happy after 30 years of being disregarded or called impractical. Or after being dissed by people who haven't studied the actual design and the practical issues that lead to it. In Ted's case there is a history of implementation attempts that have failed. Ted is a designer, not a builder, and it looks to be a source of never-ending frustration to him that he just can't go and build it himself. He appears to have built small software projects himself, but he doesn't have the implementation talent for larger ones. Taking the above a bit further, Ted is fundamentally a liberal arts guy whoses progression into computers was because he wanted to use the systems he saw as possible, and the only way was to direct it himself. *Bazzar style development wasn't available back then* as it really requires Internet availability on a large scale. He's trying the bazzar now, but the other issues and the old history get in the way. Ted also has the Cassandra problem: 20 years ago he predicted many of the problems we have with the current hypertext systems, and he designed solutions. You can guess what happened next. Ted and Doug are also getting old, and I'm certain that is a contributing factor to their attitude. They want to see somthing happen in their lifetimes, and they don't have many more chances so they have to be more forceful. This can be taken the wrong way. The good news is that Jef Raskin is young enough to continue persuing his vision over at sourceforge with a Bazzar style project. Ultimately what they want is their visions to be implemented and deliver their promised benefits. http://www.bootstrap.org/ http://sloan.stanford.edu/MouseSite/MouseSitePg1.html http://mxmora.best.vwh.net/JefRaskin.html http://humane.sourceforge.net/home/index.html http://www.sunless-sea.net/forum http://www.abora.org/ http://www.unrealities.com/robj/xancurs.htm http://xanadu.com.au/ararat http://www.nongnu.org/gzz/ Disclosure: I have met and taken a UI design course from Ted and had some fascinating discussion with him. I have never met Jef or Doug. --john jdougan@(reverse)mca.org "Computer interface design is a branch of movie making." --Theodor Holm Nelson