Re: Leventhal's challenge misses the point (long)

Subject: Re: Leventhal's challenge misses the point (long)
From: David LeBlanc <whisper@xxxxxxxxxxxxx>
Date: Thu, 27 May 1999 13:10:01 -0700
Having just looked over Mr. Leventhal's article, I had a few comments:

"XSL has nothing to offer the web"
"Page composition will not even occur on the Web as it is a notoriously
            computationally-intensive process."

Hmm... they are called webPAGES aren't they?

"Frankly, people that design page layout processes will have no difficulty
creating a CSS stylesheet for the Web and an DSSSL stylesheet for paper,
one of the solutions which is already in place today."  

DSSSL - a nightmare of complexity for both implementors and users. And,
unless something has changed, nobody has yet implemented all of DSSSL.

"A more important point is that XSL, as a web page style language, is an
enormous unknown in terms of processing efficiency." ... "These things are
not known for XSL and good professional opinion says it is going to be a dog."

I wonder if it's unknown to those early implementors of XSL? Who are these
people offering their "good professional opinion"? This argument is
reminiscent of the halabaloo  surrounding OO when it was the new kid on the
block. I don't know how it will be, so I don't want it?

"Other than printed page layout, XSL has very little to add to CSS
formatting capabilities,<snip>"

True (maybe) ... as far as it goes.. should we forget that XSL also offers
the ability to manipulate document structure which CSS is inherently
incapable of?

"The fact that interests us here however is that there already exists a way
to do tree transformation that adheres to a W3C Recommendation today. It is
called the Document Object Model (DOM) <snip>"

Why use a screwdriver to screw a screw when a hammer is available? XSL is
for transforming a document and presenting it... DOM is orientated more
towards extracting information from a document as I understand it. Using
DOM to do transformations for presentation is ah-hoc compared to using a
tool designed for the purpose.

"XSL Does Not Support Interactive Web Documents"

My coffee maker doesn't make toast either. Seriously, why do XSL and CSS
have to be regarded as competitors? It seems entirely feasible to me to
have an XSL stylesheet that takes XML document content and transforms it
into a CSS annotated webpage suitable for interaction. Along the way, I
could use the XSL stylesheet to create (for example) an (interactive) table
of contents  - maybe even in a separate frame to facilitate navigation of
my website. CSS is entirely incapable of this.

I would also point to UIML as something (not exactly related to XSL) that
CSS is not capable of either.

"Semantic Information Threatened by XSL"

A fascinating argument whose logical extension is to stay with HTML on Lynx
IMHO. He points out that FO's as the output of XSL means that a rendering
engine would not have any of the document's original semantic
information... darned if I want my printer or browser to understand the
semantics of (for example) poetry. I want it to print/display it!

"XSL is an Ugly, Difficult Language"
"XSL is DSSSL in sheep's clothing, <snip>"

U oh.. we almost agree on something. DSSSL is a BUFF, and I wonder how Mr
Leventhal can decry it here while proposing that print layout artists use
it earlier? 

XSL syntax is a bit awkward, especially if you don't think like it does. I
think using XSL requires the same sort of perceptual shift that procedural
programmers undergo when they transisition to object oriented  programming.

And, please don't make too many comparisons between XSL and DSSSL. Their
similarity is a necessary consequence of their being concerned with the
same problem domain (and I agree that too much lispishness has crept in) -
but if you suggest they're too much alike, some parenthetical person might
suggest using lisp as the scripting language for XSL (one of the things
that I truly hate about DSSSL)!

(He does make a nice argument for including a scripting tag in XSL, but
that's old chewed up ground and I'll refrain. :-) )

"XSL has Set Back the Web at least 2 Years"

Well, in web time, that's about 2 months, so no great loss.. if true. 


I somehow get the feeling that Mr Leventhal feels threatened by XSL and
other emerging standards. He seems to be making an argument that programers
ought to be the ones to do transformations and a lot of other stuff. I
don't think that supporting users creating XSL stylesheets or schemas or UI
layouts is going to put programmers, database guys or GUI developers out of
business anytime soon. If, as I infer from the tenor of his article, old
ways are better, might I suggest he retire to clay tablets and cuneiform...
if not cave walls and colored clay.

Dave LeBlanc

 XSL-List info and archive:

Current Thread