Top > Dave's World > Weblog Archive > 2000 > May > 20Previous/Next

Scripting News, the weblog started in 1997 that bootstrapped the blogging revolution.
Permanent link to archive for Saturday, May 20, 2000. Saturday, May 20, 2000

Mongolian Throat Singers

I didn't take that many pictures today, but I got some good ones.

Outside the Rijksmuseaum, I came across a group of what I thought were Tuvan throat singers. Well, it turned out they aren't Tuvan, they're Mongolian, but they're really good. A British woman said it sounded like Irish folk music, I said it sounded like American hillbilly stuff. Totally uncanny. These guys would be popular anywhere. I bought their CD, 35 guilders (I guess that's not too cheap) but why not? I already forget what the music sounds like, but I don't forget how much I liked it. I'll get to listen to it when I get back home.

Scientific American: The Throat Singers of Tuva.

(The name of the group is Altai-Hangai, which in Mongolian means Gone with the Wind.)

Amsterdam's red light district

I bet these pictures get the highest clicks!

Behind Dam Square, on a canal, is Amsterdam's world famous red light district. It's very beautiful, if it were in an American city it would be the most desirable place to live (if it weren't for the prostitution).

In Amsterdam it's done in the open.

One more pic

This is just a random canal on a gorgeous spring Saturday. Ain't it a pretty city? Don't you just want to take it home with you??

Point of view

Two interesting shifts in viewpoint, as a result of this trip. First, think of the Web as an application, or a computer, like VisiCalc, Lotus 1-2-3, or the Apple II. The "Semantic Web" is Tim Berners-Lee's version 2.0, perhaps like Symphony or the Apple III. I realized this while I was listening to Tim's conversation with the attendees of Developer's Day at lunch on Friday. As luck would have it, I got the microphone last, and said several things. First, I said the two biggest problems for the Web are patents and Microsoft's dominance of browsers.

Then I talked to Tim about the version 2.0 thing. I said he should be incredibly proud of his accomplishment, and be careful in thinking about the next level Web. It's got the largest installed base ever in the history of software. If there is to be a version 2.0, it must overcome this huge disadvantage.

Ask Microsoft how this works, their biggest competition for Windows is older versions of Windows. Upgrades, esp massive ones like the poorly described Semantic Web, are notoriously difficult to pull off. Meanwhile Microsoft is doing what they do so well, the incremental upgrade. Slowly but surely MSIE is being reconceived as an application platform. I believe this is the version 2.0, and with better search engines, and incremental additions of XML content to the HTML, and Rohit's pleasure button (more traffic if you use the XML additions) we can get Tim what he wants. But it cannot be the upheaval that the Web was because of the installed base. It's hard to appreciate how much has been invested in HTML, dirty as it is, over the last ten years.

Another shift in viewpoint. Frank Leahy of Wired gave a really interesting talk about the architecture of HotBot. Each HotBot request can result in several requests, over HTTP, using various formats, to a variety of infomediaries, including Inktomi. As he was describing this, I visualized a computer on Wired's LAN on Third Street in San Francisco, talking to a computer on Inktomi's LAN at their offices in the East Bay. (Emeryville?) I was astounded to think that a high-flow server like HotBot (which runs on NT) could make so many HTTP requests for each search and still serve 5 million or so searches a day. I think of an HTTP request as being a big expensive thing.

Well, as it turns out I had visualized it incorrectly. In fact the two computers are in the same cage at Exodus. And there's the change in view, a loop back (for me) to 1982, when I was an active Compuserve user and wanted to write software that ran on their systems. I had a BBS for the Apple II, written in Pascal, that I believed could compete favorably with the BBS software they had running on their old DEC mainframes. But no matter how I phrased the question, I don't think they ever understood what I wanted to do.

Today Exodus, and other big co-location services are doing what Compuserve might have done. I believe as the net shakes out more and more stuff will gravitate upstream. Our challenge is to do that without sacrificing the power of running your own server, which is something we plan to explore with Pike and its descendants.

Looping back to Frank's talk, he showed us how they are working with vendors to create a common XML-based syntax for the services they use to run HotBot. I asked if he planned to make this spec public and he said yes, and I said I'd like a shot at creating a Frontier-based interface for that, and he said of course. (That's how you hit my pleasure button. )

"Find a shared vision"

Once again I feel like Forrest Gump.

In Davos I asked Bill Gates if he could sign on to Bill Clinton's challenge to find a shared vision. I hoped he would say "The Web the Web, that's our shared vision!" but he didn't.

I have the same wish for Tim Berners-Lee. Like Gates, he's surrounded with super smart people who think like he does. And like Steve Jobs, Mitch Kapor, Dan Bricklin, all of whom created seminal products in a line leading to the Web, he yearns to do it again.

We're definitely at a crossroads. On June 1, we'll hear, in some detail, what Bill Gates's next vision is. Like Berners-Lee's vision it must include the Web, but it's more than the Web. In the "more" is presumably Gate's next barrier to entry. In normal times it wouldn't be as dramatic as now, because the verdict in the anti-trust trial dramatizes everything Gates does until it's overturned or implemented. Applications surely figure heavily in the next configuration of Windows, and so does the Web. Will they be Microsoft applications or will there be a framework for others' applications?

To me, the only sharable vision is the Web with nothing extra. A commitment by Microsoft to gently upgrade the Web, while competition catches up. We must have competition in Web browsers. This is something that the DOJ got right. At minimum, if Microsoft wants the respect of the people who create and use the Web, they must cooperate here, if they want to be taken in any way to be a leader, and a supporter of innovation.

On the flip side, for the W3C, I hope they too will embrace the Web as it is today, with all its warts and blemishes, for the scarred battlefield that it is, and let's add some features that press the pleasure button for the Web creators.

And to the creators (I think of Zeldman), instead of looking back, look forward. It's true that the Web is broken, but for a broken system it works remarkably well. Laugh and sing "it's even worse than it appears." Now what would you die for? What single feature would unlock a years worth of creativity for you. The more elegantly and simply you can express that, the more likely it will excite the people who create the technology that controls the Web. (MS and W3C share control of the Web, imho.)

One more thing. I plan to download and start using Amaya. I saw a demo at WWW9 and it's nice! Very different from any browser I've seen. Why didn't I download it before? Why don't we ever talk about Amaya? I have an idea that it could be made to work really well with Manila sites.


Thanks to Wes Felter, a pointer to a thread on me and SOAP and patents and RPCs and Sun on the FoRK mail list, which I'm going to join as soon as I find subscription instructions. (It's early in the morning here and I've had no coffee, usual problem!)

Apparently I misunderstood Sun's position as expressed by Anne Manes. In my defense I said I was confused. In a follow-up conversation on Thursday I expressed a wish that Sun would do something positive, create and publish a spec that it would like others to support for connecting web-based applications. I think that's where I'd like to leave it, except to say this.

We've seen so many markets turn into battlefields, my hope is that once we could have a market that involved many if not all of the big names in technology that wasn't a bloody battle.

One of the posters on the FoRK list asked why Sun is so important, that's my answer. Do computers have to be so political and bloody? I don't think they do, and the choice is in the hands of just a few people.

BTW, FoRK stands for Friends of Rohit Khare.


Last update: Saturday, May 20, 2000 at 3:33 PM Eastern.

Dave Winer Mailto icon

Click here to view the OPML version of Scripting News.

Morning Coffee Notes, an occasional podcast by Scripting News Editor, Dave Winer.

May 2000
Apr   Jun

Click here to see an XML representation of the content of this weblog.


© Copyright 1997-2005 Dave Winer. The picture at the top of the page may change from time to time. Previous graphics are archived.