Interview With Robert Cailliau - Joint Developer of the Web
- Helen Ashman:
- Welcome and thanks for sparing the time to talk with us. Firstly, can I ask if you're a regular participant in the SIGWEB (Hypertext, the Web Conference, etc) Community?
- Robert Cailliau:
- I attended the Hypertext conferences in the early 1990s. Then in 1994 I set up "my own" series by organizing the first WWW conference at CERN in May. That led to the incorporation of the first web organization, the "IW3C2" (http://www.iw3c2.org) (what a bad name) and our conference has had 13 issues in 10 years, it is still going strong. ACM is now one of its sponsors. But in 2002 I decided it was time for a change, and I left the Executive of the IW3C2 this May.
- Helen:
- What is your primary research interest and what is it about this area that drives your imagination?
- Robert:
- If you mean doing work that leads to published papers, then I have not done any research since the end of the controls project of the CERN proton synchrotron in 1980. I'm really an engineer who loves building things. Theoretical research is not what drives me. I have a continued strong interest in human-computer interaction though. It continues to amaze me how bad computers are, when there is so much we know we could improve. And I'm not even touching the web... Yet, there is also the little-recognised problem of internal behavioural complexity of computers: people can't drive them like they drive cars whatever we like to think. We will never be able to make computers that easy because they have too much internal state and our organic brains cannot cope with that. You really do have to learn a lot, no matter how good the interface. The difference between driving a car and using a computer is more like the difference between being able to read and being able to understand a text's meaning.
- Helen:
- How has your earlier work influenced both your research and technological interests? Have these influences affected your view of hypertext and the Web?
- Robert:
- In retrospect I think I have always been fascinated by the interplay of syntax and semantics. Even when I worked at the Laboratory for Mechanical Engineering I was constantly tinkering with this: I completely overhauled the lab's ways of measuring (this was 1970) and put digital readouts and computing where before there had been analog tools wielded by humans. I got into computer languages not long after, then computer-aided document writing, then hypertext (but only as an amateur). While I have written commercially successful text manipulation systems, wysiwyg editors and the like, I have actually never deeply studied the theory that was being written up at the same time in the academic world. I'm rather embarrassed by that. However, I don't feel this prevents me from believing that the early HTML was highly flawed, even very bad. Just think about it: no clear ways of nesting constructs... But it was not my fault. I wanted a clear separation of syntax and semantics, much more something like what Grif had: structural markup plus a programming language to express presentation. However, the rest of the team had a strong taboo against programming languages at the time.
- Helen:
- Is the state of technology, hypertext, and the web in 2004 as you envisioned when you started working on the Web way back in 1990?
- Robert:
- Well, to put it bluntly, I was not always on the same wavelength as the others, back in 1990-1993. I was a minority of one, but I still believed in the essence of the WWW undertaking (I don't use the word "project", because the web was never an official project at CERN, it was tolerated as something that had to prove its worth and might one day do so. A few people, notably Mike Sendall, kept supporting it). I was very taken aback though by the quite harsh attacks on the web by Ted Nelson at the 1993 Hypertext conference in Seattle. I had my own criticisms, but nothing this precise and hard. Since then, I have slowly been convinced by some of his arguments. The web has certainly not gone where we aimed it in 1990-1992. It is getting better very slowly, and I would say that most of the lagging in the adoption of standards was certainly caused by resistance from industry. But there were also some quite bad episodes of "not invented here" in the team's attitudes. Some arrogance also, which has given us some rather horrible results. Finally, the constant frictions with industrial and commercial interests often caused a great delay in providing much needed implementations of simple standards. On the other hand the web community as a whole sometimes ignored user's needs: the taboo against programming languages led to a vacuum that was filled by Javascript. I have written compilers and designed languages, but nothing I have ever seen anywhere in the whole known universe is as ugly as the Javascript+DOM combination.
- Helen:
- What would be the key thing you would change with regard to the state of technology today?
- Robert:
- Perhaps dismantle the whole Internet and Web and start looking after people's real needs? No, even I can't live without Google (or is it Teoma now?). I'd like to see a true, world-wide micropayment scheme for information. My keynote speaker at the first web conference spoke about digital cash. We're ten years later and nothing has happened. Micropayments would allow us to break the (very) vicious triangle of author-reader-sponsor. I know at least one academic who has argued (mostly from theory and observations made in the US) that people do not want micropayments. Yet I live in a country with the longest and most successful micropayment information market. I grin.
- Helen:
- If you had the choice, would you dis-invent any technological advancement?
- Robert:
- Now let's get this straight: there is knowledge, which you gain by
experiment, by observation, by building theories. You cannot and
should not stop people from knowing more or wanting to know more. As
humans we need to know more: how the climate of the planet works,
how to make the economy work for people instead of the stock market,
maybe even learn about the Higgs boson. Knowledge by itself is
neither good nor bad, only potentially useful. But then there are
applications of knowledge: technology. Almost every application has
good and bad sides. We should use technology selectively. It's not
because we know how to do something that we should also actually do
it. We seem to have learned not to use nuclear weapons. We will
need to know how to clone humans, but we should perhaps use that
knowledge only very exceptionally. We must be very very careful with
genetic engineering: remember that the rabbits in Australia were not
even engineered.
Lumping together of "science-and-technology" is a bad habit of the press, politicians and even scientists. Yes, I know we need new technology to do the next scientific experiment, but not all technology gained from science should be applied in commercial ways. I have grave doubts about the commonplace application of the computing grid and of the semantic web, however useful those technologies may be for science.
So I don't want to dis-invent anything ever, that is the wrong question. I do want to stop some applications and I would like to be very careful with large-scale use of quite a few others. We are now well past the times when you could understand a new invention by screwing the body panels off to inspect how the steam acted on the mechanical parts of your engine. We're tinkering with technologies so complex that only a few have any clue how they work, let alone foresee what the effects could be.
- Helen:
- What research are you pursuing at present?
- Robert:
- My present project is to retire by the end of the year. I'm frantically looking for a place which might be relatively safe and stable for the next few decades. I have not found anywhere reasonable yet.
- Helen:
- How do you think the Web will evolve in the future?
- Robert:
- The obvious answer is "more of this, faster that, better such-and-so". The real answer is impossible to give. It will all depend on whether legislators will be able to construct a global framework for regulating behaviour. Not content, mind you. I have absolutely nothing against "adult" content for example. But there is at least one human-relations rule that seems evident to me: whatever an author publishes to a reader should not do violence to a non-consenting third party. That rule prohibits child pornography, racist sites, libel and many other clearly undesirable sites. The rule needs to be refined, because you do want to allow certain criticism etc., but I think we know what is meant here and we will be able to produce some definitions. But, just like in traffic, it is control of behaviour that is important. For a content oriented example: you don't want your personal website to be shut down because your ISP does not like it that you wrote against religion. For a behaviour example: you do want the e-shop's purchase form to have the tick-boxes preset so that you don't get trapped into agreeing to unwanted services. We know the technology, we have no global legal framework, and I'm afraid that the two largest obstacles to getting a reasonable world-wide set of laws at this time are the USA and China. And I really am afraid of the semantic web.
- Helen:
- What one thing would you most like to see in the Web?
- Robert:
- Secure micropayments.
- Helen:
- You have been involved in the WWW conferences since the beginning. Do you feel that the conferences are still getting bigger and better?
- Robert:
- They are going along, but I have not attended the latest few. We have made several attempts to attract other communities than the web developers and academics. Somehow this has not worked. I wonder why this is so difficult, if you have any thoughts: send them to www.iw3c2.org...
- Helen:
- You have given quite a number of criticisms, is there anything good in the web?
- Robert:
- Definitely. First, there are many communities that have been able to
form only thanks to the web. These communities consist of people too
sparsely spread around the globe that they could form any
associations by chance meetings. This struck me first when
synaesthetes got in touch with me after I published my colour
alphabet. Synaesthesia is a condition rare enough that I have never
met another synaesthete. This "sparse community" effect holds for
many significant and interesting cases, perhaps the most important
one is people with rare diseases who can now exchange experiences,
make medical science progress, and above all find comfort.
Second, there is education and open access to knowledge. This is the most widely published positive effect of the web, and I will not dwell on it. One little girl in an unfavoured city neighbourhood declared in a TV interview: "The public library, yeah, they have some books, but on the web, there you find everything!"
Third, and this is my great hope: the common understanding that can flow through the web may make it into the crucial technology that will allow us to save the planet and thereby ourselves.
- Helen:
- Thankyou very much for talking to us.
Robert Cailliau was born 1947 in northern Belgium, studied Electrical and Mechanical
Engineering at the University of Ghent, became Fellow at CERN, the
European Organization for Nuclear Research in 1974. He developed
programming and documentation systems for the particle accelerator
computing systems. He became head of Office Computing Systems in 1987
and urged the study of hypertexts as a possible means of alleviating
CERN's documentation needs. He joined Tim Berners-Lee in 1990 to boost
the WWW effort. He started the WWW Conferences in 1993, helped found
the W3C in 1994 and 1995. He started the European Commission's "Web for
Schools" project. He saved the first web server (a NeXT cube) from
worse fate by exhibiting it at CERN. He is currently head of External
Communications at CERN.
M: hla on 13 Aug 2004
C: hla on 13 Aug 2004
C: hla on 13 Aug 2004