http://www.knosso.com/JMax/WWW95_report.html (PC Press Internet CD, 03/1996)These days, everybody talks about going virtual, and how the Information Highway will save us the trouble of travelling great distances just to meet people. Why, we can meet online!
But, as I've found out, nothing takes the place of a real live conference, with hundreds of like-minded individuals meeting face to face, in real time, with plenty of beer and sausage handy.
On April 6th I flew to Darmstadt, -- a town of about 140,000 people just south of Frankfurt -- home to a Technical University (TH Darmstadt), an Institute for Computer Graphics (Fraunhofer IGD), and the Third International WorldWide Web Conference. I had been accepted to be part of the volunteer staff for the conference, and thus to get a very close look at the event.
WWW '95 in Darmstadt was hosted by Fraunhofer IGD, with organizational assistance from the University of Illinois' NCSA, and corporate support from Silicon Graphics, Sun Microsystems, and a host of other Internet-relevant companies. About 1400 people came to Darmstadt for the conference, making this the largest WWW conference to date. Attendees came from 37 countries, with Germany and the U.S. being the best represented.
The conference's subtitle, "Technology, Tools, and Applications," gives a sense of the overall focus, on applications and developments and not so much on products or marketing. Interestingly, Netscape's presence in Darmstadt was almost non-existent -- possibly due to the academic leanings of the Conference Committee.
The key themes running through the conference were:
The week was divided into three parts: Tutorials and Workshops on Monday, Panels and Presentations (the conference itself) Tuesday through Thursday, and Developers Day on Friday. Monday's sessions looked at specific applications of WWW technology: kiosk systems, biology on the Web, authentication strategies, and the like. Friday's Developers' Day addressed HTML and HTTP standards, and the problems of implementing them.
Interspersed among all the talk of problems and applications and solutions were the three keynote addresses, which dramatically underscored the larger issues confronting the Web today. The keynotes were delivered by high-level "industry" representatives, as opposed to people from the developer community.
The first keynote was delivered by Thomas Reardon, technical lead of Microsoft's Windows 95 Networking Group. Reardon addressed the conference not so much as a member of the WWW community, but almost as its competition. In promoting Microsoft Network and the Internet capabilities of Windows 95, Reardon seemed to put Microsoft at odds with the tradition of discourse that has been the WWW so far.
Microsoft is aggressively pursuing the online market with Windows 95, and is very serious about being the biggest and most important player. As he outlined it, Microsoft's plan is to release a WWW browser to several million consumers by bundling it with Windows itself. A sudden influx of so many new users will have profound effects on the Web, its standards and who controls them.
Reardon's speech was largely greeted with apprehension. The computer world has seen this kind of juggernaut approach from Microsoft before, and many people I talked to saw it as a threat to the idea of the Net as a happy, informed anarchy. Reardon's message seemed to be that the larger market -- his market -- would rule.
The second keynote address was delivered the following day by Silicon Graphics' Way Ting, VP in charge of their Visual Magic division. Ting's address was very different from Reardon's -- in retrospect, maybe calculatedly so. Ting steered far clear of any self-promotion, not mentioning SGI's WebForce or WebMagic products at all. Instead, he chose to sing the praises of the WWW community in the development of the VRML standard and a working browser, WebSpace, now in its alpha stage. While this certainly makes SGI look good by association, Ting was very emphatic in directing the credit to the developers and to the VRML mailing list.
The VRML demos were very impressive, and Ting had one of the developers get up on stage to field specific questions about the implementation. It was apparent that they had done a very thorough job, and much was made of the fact that VRML hadn't even been thought of a year ago.
It was all very rousing, and the comments from the audience proved that Ting's pitch had gone over very well. "Today is day zero of a whole new era!" exclaimed one enthusiast.
The third keynote speech was very different again. On the thursday afternoon, it marked the official end of the conference. The speaker was Alan Kay, listed in the program as representing Apple Computer. However, as nearly everyone in attendance knew, Kay's reputation stretches far beyond Apple to his groundbreaking work at Xerox PARC in the 1970s. It is worth noting that Kay is not known for having much to do with the WWW.
Kay began by mentioning that his presentation would have nothing to do with Apple Computer, a statement greeted by hearty applause. Kay was notably the only presenter that declined to use the computer for his overheads, preferring instead to use 35mm slides. This set the tone for his whole talk. He led the audience on a very roundabout tour through cognitive science, issues in representation and understanding, the history of computing, hypermedia, and so forth, finally coming to an unexpected climax when he made a somewhat off-hand quip about IBM's decision to use MS-DOS as the operating system for their original PC, an operating system he said "would have been rejected in the 1960s for being too unsophisticated". The audience exploded in thunderous applause.
"Good," Kay said, "I'm glad you applauded that. Because you're in exactly the same place with the Web. Can anyone tell me the Web's equivalent to MS-DOS?" After a moment of stunned silence, a brave soul yelled, "HTML!"
Kay continued, stating that the WWW community is in a dangerous position, putting so much emphasis on the development of HTML, a system with very few redeeming features. Key went on to evangelize the object-oriented philosophy, urging WWW developers to adopt an object-based approach, passing intelligent objects to flexible interpreter/browsers instead of the current practice of passing dumb bits of data to browsers overly burdened with arbitrary rules.
Kay's talk hit hard; it was sobering, and as such the perfect closing to the conference, taking the focus away from minutae and back to the bigger picture. I hovered around the stage for a few minutes after Kay's presentation, listening to the small knot of developers pepper him with questions, especially wanting to hear his opinions of Sun Microsystems' new product, HotJava, which seems to follow a similar line to what Kay encouraged, making the browser into more of an interpreter and passing programs and scripts instead of raw data.
The final day, on Good Friday, was much more low-key. Instead of being at the conference site, "Developer's Day" was held at the Maritim hotel, where many of the delegates stayed. The discussion was about the implementation of technical standards. Eric Sink from the IETF explained that HTML level 2.0 was, for all intents and purposes, fully approved. When asked about directions toward level 3, he replied that we would likely see incremental advances in the standard, beginning with a level 2.1 that includes the full spec for doing tables. After that, parts of the proposed HTML 3 spec would be added one by one.
After listening to Dave Raggett, the developer primarily responsible for the HTML 3.0 draft and the Arena HTML3 browser, I was struck by the possibility that we might never actually reach the level 3.0 standard. Raggett's work is indeed brilliant, his browser elegant and sophisticated, but a general release of HTML level 3.0 seems a long way off. In an attempt to solve the problems that plague HTML today, it sacrifices much of the simplicity that, arguably, is responsible for the explosive growth of the Web. This, combined with the fact that the standards process moves much more slowly than the market does, suggests to me that HTML will evolve incrementally, and that we won't necessarily know where it's headed until we get there.
Which is not necessarily a bad thing. The one message I took away from Darmstadt is that the WWW is bigger than HTML. It is becoming more and more diverse. I believe we are going to see a web built of a whole variety of architectures: HTML 2.x and HTML 3.x, SGML and Adobe Acrobat, VRML and HotJava, and the dozen new things that will emerge in the next dozen months. All of these systems can co-exist, and each one will find its niche. Like all change, it will be easy for those who are open and flexible, and difficult for the big and clumsy. I can hear big institutions groaning at all the uncertainty, but the net result is diversity and health.
I was impressed by the common culture and ethic of the people I met from all over the planet. I was impressed by the high level of the discussions, even the casual ones over lunch. I was impressed by how young many of the Web VIPs are. I am convinced that the WWW is in good hands, and that it will weather the doings of large corporate interests, both those that seem unfriendly, and those who sing about "community" as well.
About the Conference:
Acronyms:
Developments:
-JMax Knossopolis