CWI celebrates the career of Web pioneer and CWI researcher Steven Pemberton on 4 October 2019 with the symposium ‘Declarative Amsterdam’. After his official retirement in June of this year, Steven has continued doing research. He is, among others, well-known for his work on (X)HTML, CSS and XForms and for the inspiring lectures he gives all over the world. And he is still going strong. The symposium starts with Steven’s XForms tutorial, a Web language that can possibly speed up programming 10 times. A good reason to interview him about his life, career and vision.
Steven, could you tell us something about your youth?
If I say that I had a fairy-tale childhood, you probably imagine princesses or white stallions. Unfortunately, it was the other sort of fairy tale, including running away from home at a fairly young age. My younger brother and I grew up in a single-parent family in St. Albans, a historical city: the school I went to was founded in 948 CE, already over a thousand years old when I was there. Stephen Hawking went there too; I didn't know him, but a classmate became a colleague of his.
Where did you study and work before CWI?
I went to Sussex University on the south coast of England. My tutor, purely by luck, was Dick Grimsdale, who built the world's first transistorised computer. He was himself a tutee of Alan Turing, which makes me I suppose a grand-tutee of Turing. (laughs) My first employment was as a programmer for the Research Support Unit of the university. From there I went to a research project in (entirely coincidentally) Turing's old department in Manchester, working on research-computer number five in the line of computers Turing himself had worked on. He was long dead by then, but there were still people there who had worked with him.
That was where the CWI connection started: I was implementing Algol 68 for the computer, which came from CWI (then called Mathematisch Centrum or MC). At conferences I would meet lots of Dutch people.
I then became a lecturer at Brighton University, where I taught, did research, wrote a book, and went to symposia (including at the MC), until out of the blue I got a phone call from the MC: could I lead a programming-language research group during the group leader's sabbatical. That was 1982. A year became two. Two became four. Four became eight. I have been here ever since.
What did you notice most when you came to Amsterdam? (I remember your popular Internet Guide to Amsterdam.)
Being here for a year (as I then thought) was for me a great opportunity to learn another language. I had always wanted to learn Dutch, knowing that it was the second-closest living language to English, so I threw myself into it, two three hours evenings per week, and a home-study course. After three months, I could speak reasonable Dutch, but the Dutch were terrible! No one would talk Dutch to me once they heard my English accent, and so there would be these infuriating, hilarious, scenes where I would be at the baker's asking in Dutch for a loaf of bread, and only getting replies in English.
As for work, what are you most proud of?
My name is attached to several things people have heard of: I co-designed ABC, that was the basis of Python, I helped design CSS, and HTML, and led the HTML work at W3C for a decade. But the work that was most satisfying for me is less known, the Views project (a browser avant la lettre), that alas got closed before it had time to fully mature. I think few people properly understood what we were trying to achieve, or what was unique about it, this being in the pre-internet, pre-web days, but in the short time the project ran we discovered so many new and exciting things that it has formed the basis of everything I have done since. It was such a rich vein that even now I haven't finished mining it.
How did you see CWI?
I had been a fan of the MC since my teenage years, and so coming to the MC was a dream come true. When I arrived it was still -- just -- MC, and still an independent institute. It seems to me that it was more free-wheeling then than now, and people did stuff because it seemed interesting and important, and not because of a national research agenda. In fact, I think some of the CWI's most interesting and important results came from that attitude (like the internet, the Web, and Python).
I see the role of the CWI to plant seeds, allowing others to be the gardeners. For instance, Lambert Meertens planted a seed to create a new, interpreted, programming language at a time when computers weren't fast enough for that, knowing that they would one day be. Guido van Rossum then cultivated that into a programming language that is now amongst the world's most-used languages.
What motivates you?
When I started as a programmer, there were little to no formal computer science studies at universities, so I did a certification at the British Computer Society, which included a project. I wrote a program (in Algol 68) that analysed grammars and produced data structures for the parser Dijkstra had invented for Algol 60. That algorithm had been formalised, and I discovered that the restrictions imposed were in fact too restrictive: a broader class of grammars were possible. That discovery gave me an incredible kick: I had discovered something new!
Although that was my initial spark, part of my later motivation has been making a difference. One of the joys of working at the CWI has been the wide range of things I could work on. For instance when we started CSS (I chaired the first style-sheets meeting), there was no guarantee that it would ever take off, let alone be a success; it has been an enormous kick to walk in to a bookshop and see a score of different books written about something I helped create.
What what do you think is in the future for the internet?
While I wouldn’t say that the internet is in its infancy, I would definitely say it is in its adolescence, with all that that implies. I personally think that HTML5 was a wrong turn that needs to be corrected. In a talk I give called “The 100 Year Web” I show how the recent developments in the web are not conducive to long-term information archival, and the actions we need to take to correct that. The web is nearly 30 years old, and any webpage from 30 years ago is still visible in any browser; I believe that that won’t be the case for many HTML5 pages being produced now.
What are your plans for the future?
I have a list of books I want to write, and I will be continuing with lots of my current work, like XForms and the Internet of Things. I have a string of lectures lined up, in London, Yerevan, Tbilisi, Amsterdam, and Bristol. Thanks to an arrangement with CWI, I'll continue writing papers, going to conferences, using my email address, updating my website, and taking advantage of a flex-desk at CWI. Apart from that, I'll continue singing and I’m also going to try arranging some music for choir. And tomorrow, the day after my farewell symposium, I will give a lecture during CWI's Open Day at the Amsterdam Science Park. Everybody is welcome.
Thanks a lot, Steven!