Seeing as it's the 10th Anniversary of XML, I thought I'd reminisce a little about my small involvement in its beginnings.
In 1994, I was interested in two seemingly distinct fields—corpus linguistics and Web publishing—that happened to both have something to do with SGML. So I decided to learn more about SGML.
I rapidly became an SGML snob of sorts, lamenting the state of HTML and arguing on various mailing lists and newsgroups how much better the Web would be with SGML.
Towards the end of 1995, I started to think seriously about some of the things that would be necessary to make SGML on the Web a reality. One component, I thought, would be a way of resolving SGML's formal public identifiers. SGML Open (which became OASIS eventually) had catalogs but they were like hosts files. What was needed was something more like DNS.
So I came up with an extension to SGML Open Catalogs to allow for hierarchical resolution. You can read a version of my proposal at: http://jtauber.com/1996/03/standards/fpi-urn/delegate.html
This got the attention of Eduardo Gutentag at Sun who thought Jon Bosak (who had recently joined Sun from Novell) might be interested. It also got the attention of Paul Grosso who invited me to present the idea to a meeting of SGML Open in early 1996. I was interning at Sun Labs in Chelmsford at the time and Sun agreed to fly me over to Long Beach for the meeting. My idea seemed to be received well—Charles Goldfarb described it as a "dandy" which was very exciting for a 22-year-old SGML aspirant.
I spent some time after the meeting talking to Jon and Eduardo about SGML on the Web.
After my internship, Sun offered me a job (actually a couple of different positions, including running a website for reusable Java classes, but that's another story). I also interviewed with EBT. For personal reasons, however, I decided to move back to Australia.
I was still interested in SGML on the Web, though, and the next component I thought would be necessary was an implementation of the DSSSL stylesheet language. My vision was to build a tool that would output the result of DSSSL to both Tk widgets for online display and something like PDF for printing.
A couple of months after my return to Australia, I was emailing with Jon Bosak about my DSSSL ideas. And that's when he told me some amazing news: he had convinced the W3C to let him set up a group to look at SGML on the Web and he asked if I'd be interested in being part of the group.
(to be continued)