Gorillas in the mist
Despite Sun’s pivotal role in Web services development, Microsoft and IBM are still muscling their rival away from WS-I
See correction below
JON BOSAK HATES smoke-filled rooms. Well, hate may not be the right word. Jon actually said: “I have a strong antipathy towards smoke-filled rooms. I find that whole concept repellent. And I have to believe — and maybe I’m kidding myself, but I have to believe — that eventually there is some kind of karma to that, that comes around and bites you.”
Jon wanted us to know he was speaking not as a distinguished engineer in Sun’s XML Technology Center, but personally. The occasion was a round table of InfoWorld editors and Sun standards experts, including Chief Technology Evangelist Simon Phipps, Standards Coordinator Simon Nicholson, and XML and Java Product Manager Ed Julson. And the subject of Jon’s comment was the IBM-and Microsoft-led WS-I (Web Services Interoperability Organization).
In the wake of IBM’s and Microsoft’s decision to exclude Sun from WS-I founding-member status, Sun asked the reluctant Bosak to take a more visible role at the JavaOne developer conference. Faced with more and more questions about Sun’s credibility as a player in the Web services arena, Sun officials from Scott McNealy to Pat Sueltz regularly mention Bosak’s seminal role as a leader of the XML movement.
Bosak described how XML grew out of the intersection of SGML (Standard Generalized Markup Language) and the Web, as represented by HTML, itself an SGML application. “It took a while for the SGML folks to wake up to the fact that there was something really interesting happening here rather than just a simplistic SGML application. And there were a lot of efforts to try to put the two communities together.”
“That was the mantra — let’s get SGML on the Web,” Bosak continued. “And finally [we] succeeded in getting the W3C to say ‘well, if you want it on the Web so badly, why doesn’t Sun’ — for whom I had just gone to work — ‘why don’t they fund the creation of a working group to do that?’ ”
At SGML Europe, Bosak got his boss, Mike Rogers, to authorize the project. “Neither one of us went deeply into our reasons for doing this. It was like: SGML, we know SGML, SGML good. Web good. Web need SGML. Need to happen. It was that [simple], all of about three seconds.”
Bosak talked 12 of the world’s experts into participating, including James Clark, Makato Morata of Fuji Xerox, Tim Bray, Yyves Mailer, Dan Connolly from the W3C (World Wide Web Consortium), Jean Paoli of Microsoft, and others. For two years beginning in 1998, Bosak chaired the coordination group that split up the work among five working groups. Out of these groups came XPointer, XPath, XSL, namespaces, and eventually, XML Schemas: in short, the XML foundation technologies without which there would be no Web services.
Flash forward, and the standards process now appears under attack. “One of the things that made standards processes work,” Julson said, “is they’ve been perceived as a level playing field and not representing certain vendors’ agendas.”
Nicholson rejects shortcuts around groups such as the W3C. “They have a track record of delivering, time and time and time again, technologies that have been through an appropriate process of due diligence, and have gone through a standards process that we all buy into and we all subscribe to.”
“Now, that takes time,” Nicholson continued, “There is no secret sauce there. Fine, you have a new organization, call it whatever you like, but there’s no way of short-circuiting the fact that you have to get individuals around the table like this, get some issues on the table, get them openly discussed and reviewed, work through the issues that arise out of that, come to a consensus of opinion, and then go through the publication process.”
“We’ve just seen a real downhill slide over the last five years,” Bosak said, shaking his head. “One of the really remarkable things about the XML working group then is that we all checked our company affiliations at the door. … We had the only two SGML companies who mattered, who were bitter rivals with each other — they worked together, like twins. We had Microsoft and Sun working together without any regard, basically, for competitive jockeying.”
Today we have WS-I, which Bosak endorses conceptually. “The idea of getting some interoperability going is a good idea and I can’t fault that,” he acknowledged. But when Nicholson identified WS-I as an interoperability forum rather than a standards group and I suggested that it has the function of a standards body, Bosak agreed: “From a marketing or political viewpoint, it’s going to have that effect.”
Bosak summarized his position succinctly: “Any real standards process is going to operate much more slowly than the way Microsoft, et al., are going about this, period. End of story.” Simon the Evangelist jumped in: “What we are not doing is devoting the same resources to marketing staff who call themselves standards people.”
“And consequently, when there are purely political discussions going on, you often find that Sun doesn’t have a spokesperson there because Sun is actually doing the work rather than shaping the politics,” Phipps continued. “It seems that one of our competitors has come up with a cunning ploy to exploit our resistance to doing standards hype and use it to marginalize us.”
“Who is that?” I asked. “Microsoft?” “No,” Phipps replied. “Microsoft [is] not playing the game that I just described. On the whole, Microsoft’s involvement in XML, at a technical level, tends to be very pure, constructive, and good.”
“So you’re talking about IBM?” I asked. “I’m talking about IBM,” Phipps replied. “IBM devotes more time to the politics than it does to the technology, in most cases.”
More than a month has passed since gorillas IBM and Microsoft excluded Sun from WS-I. News reports suggest they continue to veto Sun’s request to join the founding board. Why won’t Bob Sutor of IBM fix this bug? Time’s a-wastin’.
Correction
In this article, we misspelled the names of Eve Maler and Murata Makoto, two of the collaborators in the W3C’s originating XML effort.
