Rant : "Microsoft is compliant with the XSL spec"

Subject: Rant : "Microsoft is compliant with the XSL spec"
From: Warren Hedley <w.hedley@xxxxxxxxxxxxxx>
Date: Wed, 12 Apr 2000 22:09:36 +1200
I thought I'd take the time to contribute my M$ rant to the XSL
list - hopefully someone in the great monopoly may even take
notice (yeah right.) Apologies to Jonathan Marsh and colleagues -
this is not aimed at you - but at the M$ marketing machine and
M$'s priorities.

I recently had the pleasure of attending the "XML Action" seminar
in Auckland, New Zealand, where I was going to find out about

"... essential XML topics, such as the XML Data Model, XML Object Model
(DOM), and XML Data Schemas, as well as techniques for using XML on
the client and on the server will be presented. Finally, the Microsoft
initiatives and current implementations for XML technologies will be
demonstrated, and illustrated with code examples."


(I think the page at that URL has changed - it originally mentioned
XSL and its uses specifically).

In the first session I sat through a presentation on "cross-platform"
DOM where you can call methods like loadXML(), selectNodes(), and
of course transformNode(). I was shown cool things like embedding
XML data islands with <xml> tags ("no more <object> tags"), and
table-binding (although we were later shown how to construct the same
table with XSLT). When I asked, I was assured that this conformed to
the HTML spec. This, after showing an HTML 4 document with no DOCTYPE
declaration, and a <script> element after the <html> element was closed.
I'm reasonably sure I recognised some examples that I received with
the Datachannel parser (hands up who found a bug in that) over a year

Between the sessions, anticipating a similar treatment to "XSL", I
approached the presenter and asked him if he'd mind stating, during
the upcoming session, which parts were pure XSLT and which were
M$XSL. I was assured that he'd only be presenting things that
conformed to the "draft". I mentioned that XSLT had been a
recommendation for about 6 months, which was news to him, and also
told him there had been a lot of discussion on this esteemed list
about why M$ has largely ignored the spec, but is happy to implement
their own extensions first. Understandably, I was received somewhat

Anyway, the "XSL" session got underway with a few simple examples, all
of which had the wrong namespace. And sure, many of the XML documents
had no <?xml ...?> declaration. I could live with that, most of the
people there probably aren't at the bleeding edge, like this list's
suscribers. Then we got on to the more complicated examples and things
  <xsl:define-template-set> and
  <xsl:for-each order-by="+category;+title">
started popping up. The presenter mentioned that, indeed, there had
been some debate about M$'s compliance to the spec (impressive memory)
but as far as he knew "Microsoft is compliant with the XSL spec". It
was about at this point, that I buried my head in my hands and stopped
taking notes.

Actually the highlight of the seminar was when someone asked what the
difference between XML-Schema and DTDs was. He explained this well. The
next question was, where is the specification for DTDs? The answer: "I
don't know."

I was reasonably happy to sit back and not interrupt the seminar - it
would be unfair to the everyone else. But it really irks me that someone
can present so much disinformation to 80-100 keen XML developers in one
sitting. Most people can cope with a few errors in the examples, but
being told at every step of the way that everything is "cross-platform"
or "cross-browser" or "compliant with the spec", which was 90% of the
time just plain wrong, really rips my nightie (colloquial NZ - you get
the idea).

What can one actually do to make a difference?

I cornered the presenter again afterwards and handed him a scribbled list
of some of his mistakes, which he accepted quite graciously. That's all
I've been able to think of so far. I'm now considering forwarding this
message to him.

Well, I feel better now, having shared that with someone. Comments and
pro-M$ flames welcome. 

Warren Hedley
Department of Engineering Science
Auckland University
New Zealand

 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list

Current Thread