In May 1994 I attended WWW’94, the inaugural World Wide Web conference at CERN, outside Geneva.
I reported on the conference on Lynx, the online magazine established by Tony Jewell and Bernard Jauregui at Cityscape, their ground-breaking ISP. Lynx, and Cityscape, are long gone but I’ve found the articles cached on the Wayback Machine and here they are:
To launch ‘The Lynx’ magazine, the latest CityScape/Global On-Line project, our first issue is a special on the WWW ’94 conference in CERN.
At no expense spared we have sent Bill Thompson, our roving reporter, to sunny Switzerland to give us the latest, up to date, inside views of what is happening at this important event.
The Lynx is an attempt to see if a real magazine can be published on the World Wide Web on a regular basis. The magazine will cover the more alternative sides of the Web and the Internet, and we hope to be amusing, informative, and slightly anarchic …
We need your articles and columns to make sure this magazine works, and in return we will carry full resumes and CVs of anyone that works with us. The magazine is run from the Global On-Line server in Cambridge, UK, one of Europes largest commercial servers, with nearly a quarter of a million access a month, so we hope to get considerable publicity for this venture.
__ ___ EMail: to…@cityscape.co.uk
/ |_ Tel: (UK) 0223 566950
\__ ityscape |__Mail
Live From WWW ’94 at CERN
Reporter: Bill Thompson Live at CERN
In This Issue …
- A note from the Editor
- A note from Bill, our roving reporter ..
- Future Developments News from the Show
- Introduction In the beginning …
- Conference Plenary
- Wednesday 25 June Afternoon Session
- Thursday 26 June Morning Session
- Thursday 26 June Afternoon Session
- PostScript: A Final Message ?
Welcome to WWW94. I am attending the conference as a delegate from Unipalm Group plc (whose Web server is here ). Since we have network connectivity from the Conference, CityScape have agreed to make my dispatches from CERN available on the Web. There is a lot more information available on the WWW94 home pages, and for those of you with access, there is an MBONE broadcast of elements of the conference.
The Web is more than a technology: it is a new medium. Old models of publishing no longer apply: these pages will grow over the next three days as I get time to write up my notes, and they will provide a record of one person’s view of the conference. I hope you find them interesting.
My email is email@example.com, and I am reading mail from the Conference, so let me know what you think.
WWW Consortium Pre-announcement Made
12:39 Wed May 25
Speaking today at the WWW94 conference, Tim Berners-Lee of CERN, the originator of the Web, pre-announced the creation of a Web consortium.
The body, as yet unnamed, will represent the Web to the world. It will be funded by a number of sources, and will include industry members.
The European Centre will be at CERN, with a US centre at MIT. Other institutes will be invited to participate.
ESPRIT money will be available (no details), and industrial partners in the consortium have been invited to join..
It is NOT all tied up yet, but this is an important step forward in the evolution of the World Wide Web, and provides for the development of WWW and its underlyinng protocols on the same lines as the X Consortium promoted the growth and eventual commercial acceptance of X.
Sitting in the crowded auditorium awaiting the start of WWW94 the sense was less of making history than being relieved to be there. We all *know* that what we’re doing is important, and don’t need this Conference to tell us. What we want from the conference is the contact, the creation of a sense of community, the translation of virtual, web-mediated relationships into genuine human contact.
I remember HCI’86, when it suddenly dawned on us all that we had a new discipline to work in: the same thing is happening here. Already we have heard that a Web consortium will be established, and stuck on the walls are notices announcing the formation of a Web Special Interest Group, SIGWEB .
Things are happening fast.
Welcome: Professor Walter Hoogland, Director of Research, CERN
Walter Hoogland’s welcome was deeply felt. He said he was “happy to see so many faces in front of me”, and we all realised that he meant it. This was not what CERN had anticipated when they established the Web, and for most of the last 12 months CERN’s public position has been that they will hand over the Web work to another, more appropriate, body, since their business is particle physics and not global distributed information systems..
And yet, now CERN is to be the lead European centre for the Web consortium, and the CERN team are acknowledged as key figures in the development of the Web. How can anyone at CERN feel anything but pride at what they have built?
There are problems, and Hoogland identified some of them: the Web-related traffic on the NSFNet backbone is doubling every 2 months, with a lot of the traffic caused be useless graphics, inappropriate information acquisition and casual browsing of the Web. These problems cost time and bandwidth, and limit the usefulness of the Web as a whole.
Finally, applications for fellowships devoted entirely to putting existing data onto the Web are now being received, and to Hoogland this is an inappropriate use of research funding, but he acknowledged the growing importance of the Web to a large number of people both inside and outside acadmemia.
Overall, Hoogland’s message was a positive one, asking us to reflect on the growth of the Web and its current nature, and to look carefully and present and future possibilities.. and problems.
Keynote Address: Effective Rules in Cyberspace.
Dr David Chaum, DigiCash David Chaum’s beard is bushy but carefully trimmed, and his long hair is carefully pulled back into a well controlled pony tail. He wears his business suit with slightly less unease than one might expect, and gazes out at the assembled WWW94 crowd through thin tortoiseshell glasses. He used to be a full-time mathematician; he knows a *lot* about cryptography; he has used the World Wide Web – and now he wants to talk to us about Digital Cash and the nature of the society we are all creating out there in Cyberspace.
Chaum’s primary motivation for coming to WWW94 is a political one: he has realised that we are in danger of letting raw capitalsm drive the evolution of the information space, and he wants to avoid this. Unfortunatly, he has a dilemma: he may not *like* capitalism but at heart he believes in the American Way, and the American Way is based on money and the free market. Also, he has a product to market – digital cash.
We all know that money can be moved over the net, and some of us have sent credit card details etc to pay for goods and services electronically. It’s like using the phone to order a pizza. Even without strong encryption, it is a developing area and initiatives like CommerceNet show how things are likely to go. However, digital credit cards are as traceable as ordinary credit cards. If we are to preserve the equivalence of the real and virtual worlds, then we really need an analogue of cash: a token of value which is inherently untraceable. Digital cash is just this: a means of generating value-holding tokens which can be exchanged for good and services but are untraceable to the user. Check out their server for details.
In order to promote this idea, which I find appealing and sensible, Chaum proposes a model of connectivity in which there are two main choices: a “head end” (analagous to the cable feed) in which a set-top box or other central funnel controls all access, versus an open network approach in which any service/information/goods provider can be accessed. The head end approach implies two things: traceability of all transactions, and control of interactions, since all suppliers must contract with the head end controller to reach the market.
Most of the paper concerned brief descriptions of the cryptographic tricks that can be used to build a cash-oriented network society. Along with some non-confrontational attacks (“I don’t mean to be critical, but…) on the public key encryption of digital signatures lobby (you know who you are…), he provided a solid insight into ways in which we could control this new culture.
The applications go beyond money to cover things like voting rights, participation in debates, verified signatures on legal documents and so on. All of the things which currently rely on physical tokens (ballot papers, signed contracts etc.) can be managed in the open networking model using cryptographic techniques. “Cryptography can empower citizenship in open networks”, as he said. It will enable us to “build the best world we can”.
Despite the poorly articulated political analysis, and his apparent discomfort with his position as advocate of free marketeering, Chaum made a good case for his approach. As he said to one questioner at the end: “do we want data fascism or a new world?”
Bill Thompson. May 25th 1994. CERN, Geneva
Commentary on Applications Stream
First day, and after lunch we have a choice of sessions. I am going along to the Applications session rather than the technology series, and giving the workshops a miss. For details of all the papers, see the WWW94 pages at CERN.
The focus of the presentations this afternoon was “publishing”, with four papers given. All were directly concerned with extending the rather limited WWW paradigm beyond forms-based interaction (the revenge of the iBM 3270 as David Wetherall of MIT put it in his paper).
In the first paper, given by Steffen Meschkat (firstname.lastname@example.org) we saw how ART+COM in Germany have constructed a first pass at an interactive journal, with a paper on a software Volume Rendererpresented as a WWW document. This means that as well as “reading” the paper I can also try out the software, download the source and even use it as a resource.
The scope for this method of publishing is immense: it challenges the print-based model by removing the distinction between authoring and publishing, and blurring the distinction between pre- and post-publication. Despite the fear of revisionism expressed by one questioner, the ability to interact with a journal article seems to offer great potential.
John Mallory of the AI Lab at MIT, Anders Klemets of the Swedish Royal Institute of Technology, and David Wetherall of the MIT Computer Science Lab all demonstrated other ways of integrating non-standard services with a HTML-based browser such as Mosaic.
Mallory’s testbed application allows the user to query a database of White House publications via a forms-based interface. The server is implemented in Common Lisp and builds the HTML it sends to the client on the fly.
In contrast to this handcrafted code approach, the system described by Klemets integrates the MBONE-conformant audio and video tools (ivs, nv, vat ) with the Web to allow a forms-based interface to start an asynchronous sound/video feed to the client system. This gets around one of the main problems with the Common Gateway Interface – the need for a called program to run to completion before the client can continue. The demonstration system allows the user to select and play a number of audio and MPEG clips.
Finally, David Wetherall described two systems based on active pages, using CGI-based programs (or Perl scripts!) to manipulate local databases. His two examples are a Web-based DNS manager and a local White Pages system. Both are working systems and use CGI programs to manage concurrency and validation at the server side. However, the batch mode limitations of the Web cause problems, and Wetherall called for a model which more closely resembled the X Protocol. Since the stateless nature of HTTP is one of its main benefits for wide area information service provision, this is unlikely to happen.
All of these systems are interesting because they move the Web away from its original data publishing metaphor towards a database management approach. What is worrying, however, is that much of the work described in the session could be done trivially with existing relational DBMS tools.
We need to ask ourselves whether this work on extending the HTML/HTTP batch-oriented approach towards a distributed database system is a sensible way forward for the Web, or whether it would be more sensible to attempt to incorporate the existing, well-understood, DBMS work in a way which complements the Web’s underlying capability.
Bill Thompson, CERN, Geneva
Yesterday’s sun has gone, and steady rainfall greeted us as we boarded the coaches from our hotel to CERN. Pieter van Brakel from Rand Afrikaans University had spent all of yesterday carrying his jacket around: now he stood in a short-sleeved shirt and shivered in the rain.
Commentary on Education Stream
The paper from Bertrand Ibrahim of the University of Geneva, entitled World-Wide Algorithm Animation, was ostensibly about the use of WWW in education, but really it was about another means of getting around the batch-oriented limitations. In this case the driving problem was to build a simple debugger interface to Pascal programs that would work with Mosaic/httpd using HTTP.
The solution used, and available in French, uses a CGI script to spawn the required program as a child process. The PID and machine name of the spawned process are passed back to the client side and can be given to the server. This allows the server to run an arbitrary CGI script which can open a pipe to the spawned process and issue instructions to it.
With this mechanism, asynchronous program execution and control are possible from within the Web. The controlled program must be written to work with this mechanism, but this is not a difficult technical issue.
Next up was Pieter van Brakel from Rand Afrikaans University, an Information Scientist who is using WWW and Mosaic to teach his students about hypertexts and information management.
His paper was a useful corrective to the technical focus of other presentations. He is concerned with the content and structure of the Web rather than its technical base. It is apparent that, while we may be seriously smart developers of startlingly original interfaces, most of us couldn’t create a usable hypertext in a hundred years.
Although he was too polite to name names, van Brakel mentioned that he shows his students good and bad examples of Web design. Tonight we have the first Web awards, and you can vote here , so we’ll see some of the good ones. As far as I know (unless CityScape want to organise it), there isn’t yet a Web turkey award.
The educational benefits of Ibrahim’s work are obvious: by giving students a forms-based interface to a debugger they are able to focus on the problem (understanding the code) rather than the tool (using sdb or dbx). But the wider implications are that the batch-oriented limitations of HTML/HTTP can be overcome in an imaginative and flexible way.
The serious point that underlies van Brakel’s paper is that too many of us who are weaving webs are techies who got sucked in: someone in the organisation had to set up the Web server, it was *you*, and now you’re the (unofficial) Web Weaver. However, the structure of the information held is not your main concern: perhaps it’s having some really neat hotlinks, or flash gifs (“You flash gif” could become a new term of abuse, I suppose), or a clever CGI script to do something on the fly. Just like the writers who got access to PageMaker on the Mac and thought they were automatically graphic designers, we are building Webs and acting like librarians. >
Bill Thompson, CERN, Geneva
Commentary on Business Stream
Lunch was nice: I got to eat today instead of just have a coffee while I typed my notes. I sat and talked with Carlye Hogin of the Science Policy Research Institute, and we talked about art and philosophy and analysis and our (separate) experiences in the bar at the ICA, London.
It was all about the Web, really: in the atmosphere of a conference you can’t help bringing everything back to the main subject. We talked about Teresa Brennan’s new book on early experience, and designed a Web-based psychoanalytic tool whic would allow you to reenact traumatic experiences in cyberspace: a virtual analysis. We also talked about the Situationist International and their approach to popular culture and popular media. If the Web is to be the forerunner of cyberspace, then someone needs to inject a bit of subversive humour and confrontational politics into it. Now seems a good time to start.
This afternoon I attended the business/press stream: intended as a way of sucking serious capitalism and media barons into our way of thinking, only six people at Robert Cailliau’s session were not already Web-initiates, and one of them was Tim Berners-Lee’s father!
Robert went through with his demonstration anyway: a useful short trip around the Web. He even risked all when asked to find the online copy of a slide he had used – and got a fulsome round of applause when he found it under here !
He was followed by Russ Jones of Digital , who gave a well-positioned description of the issues Digital faced when they decided to set up a Web server to provide access to their online product information. Since the data was already present as a well-structured ftp archive, they built their Web on top, using tools to generate document citations from the text abstracts held on the ftp site, and then constructed a range of search tools.
If you have visited the Digital server, you’ll see how clean it is. One of the important design decisions they made was to treat every page as a potential entry point: you can never force someone to come into your Web space through your designated home page, so you should not design the space as if this was happening.
Next up is Ray Anderson of IXI, who is going to talk about the ways Mosaic has changed IXI’s corporate culture…. watch this electron.
Bill Thompson, CERN, Geneva
and that was all I wrote – an example of early blogging, perhaps? Tony and Bernard loved it, lots of people read it, but not everybody liked it – though I’d point out that at the time I actually was a professional journalist – not just a hack typing from my diary!