Weighing in on XSL / Standards

[Chris Lilley]

David Megginson wrote:
> 
> Joshua E. Smith writes:
> 
> [on XSL]
> 
>  > Am I missing some history here?  When a detractor can say, with a
>  > straight face, that he doubts the standard can actually be
>  > implemented, that's usually a sign that the standard is *way* out
>  > in front of the community, and needs more baking.
> 
> XSL inherits from at least two major sources -- FOSIs, which are/were
> SGML instance-based stylesheets for typesetting SGML documents,
> specified by the USDoD; and DSSSL, which is/was a declarative
> Scheme-based language for transforming and formatting SGML documents,
> standardized by ISO.  

Uh, it also inherits rather substantially from CSS, if you look at the
drafts.


> I don't think that anyone can say, with a straight face, that XSL
> *cannot* be implemented, 

Right. They can claim that its hard; they can claim that the effort is
not worthwhile, (which depends on your particular cost/benefit analysis)
but can only claim that they *can't* be implemented by pointing to a
direct contradiction in the spec, and furthermore one which cannot be
removed.

> though the flow objects will be a challenge
> (as with CSS, for that matter) 

I don't see FOs as a particular challenge, but I know that some folk,
particularly those with a very low-level, bit twiddling view, tend to
find them scary. But then, thats probably because CS students are taught
all about complier technology but aren't taught all about
constraint-based geometric systems.

> -- the implementation process is fairly
> straight-forward, and there are already several implementations
> available (none is perfect, but that's to be expected when the spec
> itself is still unstable).  In this regard, XSL is doing much better
> than many other W3C specs.

Hmm.

Talking specifically about the FO part and about announced
implementations, I would characterise them as "fairly incomplete" rather
than "not perfect". And certainly not "much better than most W3C specs"
unless you can quote counterexamples.

> The real question is what kind of market penetration XSL will have.
> I'll probably use it a lot for my personal work, because I'm used to
> that kind of programming and quite enjoy it; you may be right, though,
> that it's too far in front of the Web community -- even a
> brain-dead-simple thing like CSS1 is still having trouble catching on
> after several years, 

Primarily because of the difficulty, at that time, of changing a world
view that said "tags and attributes are just fine, thanks, and can I
have another 5 HTML extensions".

CSS1 is not brain-dead-simple, from an implementation standpoint,
although you are welcome to whip up a quick 100% compliant
implementation if you wish to prove me wrong. 

But to agree with your earlier point, the implementation process is
indeed relatively straightforward, provided that compliance to
undocumented HTML layout features and not breaking HTML existing pages
isn't taken as the overriding design constraint.

The XSL FO implementations that I have seen are significantly behind
CSS1 in functionality, though doubtless they will catch up soon because
the main difficulty in the past was not technical but political and
sociological. People didn't see the point in style sheets, and in many
cases they still don't - but at least XML doesn't have a huge legacy
rendering issue to deal with.

> and XSL is at least two orders of magnitude more
> complex than CSS.

Apart from the addition of transformation, which is a separate spec and
can just as well be used to generate XML+CSS as it can be used to
generate FOs, I don't see a whole order of magnitude more complexity (or
functionality) in XSL than in CSS2.

--
Chris


xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev at ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ and on CD-ROM/ISBN 981-02-3594-1
To (un)subscribe, mailto:majordomo at ic.ac.uk the following message;
(un)subscribe xml-dev
To subscribe to the digests, mailto:majordomo at ic.ac.uk the following message;
subscribe xml-dev-digest
List coordinator, Henry Rzepa (mailto:rzepa at ic.ac.uk)