RE: Venting

Subject: RE: Venting
From: "Didier PH Martin" <martind@xxxxxxxxxxxxx>
Date: Thu, 4 Feb 1999 19:01:43 -0500
Hi Chris,

This is an interesting assertion.  The work on XSL was started by Jon
Bosak, who put a lot of effort into DSSSL and led the informal DSSSL-o
profile group.  All three of DSSSL's editors (Sharon Adler, Anders
Berglund, and James Clark) sit on the XSL WG; one co-chairs it, and
another edits half of the XSL spec.

How this disrespects, rather than learns from, DSSSL eludes me.

You are right these persons where part of DSSSL specs and learned a lot
about it. This is why we got the transformation part that came so fast and
indead showed that lessons where learned from DSSSL. The comment about
disrecpect is more toward the fact that W3 just created a new language
instead of improving the one already there and named DSSSL. This notheless
do not mean that XSL is not well done, at least the transformation part. No,
in fact, the transformation part is more seamlessly integrated with HTML
constructs. Even more and better than dsssl. What is however somewhat harder
to understand is a) create a style languaged named CSS and have part of XSL
compete against it. Also, as Paul and I already said, only the formatting
part of dsssl is a de facto standard because only this part got implemented
in real applications. This said, I do not and that is not in my mind that
these individual didn't learned nor did they applied their knowledge which
toward the goal of creating a style language. So, the point is that W3 as an
institution (not necessarily the above mentionned individuals) choosed to
create a new language. Thanx, the transformation part is great and resolved
some bugs we got in dsssl (just a mention here: a dsssl document is a SGML
document SGML let you choose delimiters and this is whay we have ()))
because this was choosen to be the tag delimiters and also the fact that
such DTD would be scheme compatible - it remains that we could have debugged
this and simplify the notation - James did that by introducing the macro
concept - better than that merge the two. No, the choice was made to redo.
End of the story.

This said, many thanx to all the individuals you mentionned (and a special
thanx to Jon Bosak and James Clark) for their efforts and knowledge. But it
remains that splitting the transformation part of XSL could be a good thing:

- economical reasons
- The last Microsoft patent annoncement
- You guys are defending standards and so on, this is to provide feedom of
choice to users (the right to have the freedom to choose your vendor in
fact, not the right to get innovation). This is what we want to (the right
to choose the vendor) and not get the market or power only in the hands of a
- there is momentum a least among the early investors (people who created
implementations) and early adopters (people who used it and gave feedback to
the former)
- There won't be any confusion about what is provided to the user. If we
retain XTL as a word to describe the transformation part then, at least the
end user will know that XTL is something concrete not a part of, yes
but....or wait until that big guy have it fully implemented (or maybe never
as they whish after all, they have the market, why rush?)
- This can reach the market beach at least this year and will indead fulfill
Jon's goal to have XML be more real stuff.
- I know, being in the position of power and being challenged by the
population is not easy. But this movement is based on good grounds. But this
side of the fence has an opinion too. A standard is not only owned by the W3
members, now that other people invested efforts in it, it _their_ standard.
Is it not what standards are made for? What we ask is just that, that it is
a standard owned not solely by a few and that the rewards of such standards
is not also solely reserved to a few. so don't get me wrong, I repect these
people but also repect people that brought reality to XSL and these people
are this list participant (and others collegues we don't know or do not
speak often). Period.

Didier PH Martin

 XSL-List info and archive:

Current Thread