Bray recalls team XML
Shoulda happened sooner
10th birthday interview In an industry characterized by cold logic and technology, the contributions of real people often get lost.
Publicly, great things are seen to be achieved by impersonal corporations and institutions - rarely individuals. In an effort to put this right as far as XML is concerned, Tim Bray, Sun Microsystems' director of web technologies and one of the original architects of XML, has published a detailed - and often moving - blog to celebrate the tenth birthday of XML and the people who made it happen.
Starting with pioneers such as Ted Nelson (famous for hypertext), Tim Berners Lee (he of the World Wide Web) and Charles Goldfarb (known for Standardized General Markup Language - SGML), Bray works his way through a long list of personalities who contributed to the XML cause.
He even anthropomorphizes corporations so Netscape becomes "Ned" ("a rookie pitcher who burns 'em up in the first 20 games with a wicked screwball") and Microsoft becomes "Mick" (a domineering, ruthless, greedy, egotistical, self-centered, paranoid bastard").
Bray's decision to go and work for Ned almost lost him his role as co-editor of the XML spec when Mick protested that this would give Ned too much influence. We all know how that particular innings of the great internet ball game ended. However, a compromise was worked out on XML at least that saved the day, with Microsoft's Jean Paoli - then an Internet Explorer 4.0 product manager now senior director of XML architecture - joining the XML team.
The first draft XML spec was released by November 1996. A year and couple of months later we had XML 1.0, and thusly SOAP, web services and Service Oriented Architectures (SOAs).
Ten years on Bray told Reg Dev that he is mostly pleased with what XML had achieved - although it was not necessarily what was planned.
"When you set out to do something like XML you have no real idea of what it will be used for. At one level it has been very successful and there are many areas where it has been used that we did not anticipate. We thought of it as a mechanism for professional publishing activities and hoped that a lot of HTML stuff would have migrated to it."
Bray is surprised it took so long into the story of the computer industry, 1998, for something like XML to appear. "We were quite surprised by the lack of skepticism when we started to push the idea of XML. People just accepted it. It should really have happened in the 1970s."
The parallel development of the open source software movement - that also celebrates its tenth anniversary this month - was a strong factor behind the rapid spread of XML, according to Bray.
"I think open source would have happened without XML and vice versa so they are not necessarily linked. But the availability of open source parsers for XML certainly helped. Previous attempts at data packaging standards were frustrated because all the parsers were expensive commercial products. And I guess there is a certain parallel culture between open source and XML in that both should be open and transparent."
With hindsight there are a few things in XML Bray would change - mostly to simplify it. "It's too late to change it now, but it is really too big and too complicated. We originally focused on being compatible with SGML and included a bunch of features that don't actually get used. The notion of entities as storage units and unparsed entities have not really added anything."
Bray expects the influence of XML to decline as new ways to achieve the same objective emerge. "It was the first but it doesn't mean it was the best. I think AJAX and standards like JSON are beginning to show that you can do the job better. But XML certainly laid the ground for loosely coupled, web-based services," Bray said.®
Sponsored: Hyper-scale data management