RE: DSSSList, Year Two

["Jonathan Borden" <jborden@xxxxxxxxxxxx>]

Comments on DSSSL:

	The 'problem' with DSSSL, in my view, is that the field has been
concentrated primarily on formatting output and not on transformation. XSL
has risen in popularity because of its current concentration on
transformation (there's hardly any formatting work going on at present).
Simple stuff is fine in XSL but as I've attempted to implement more complex
stuff, I find a real need to create XSL constructs analagous to lambda,
defun etc. Once you start to do the equivalent of dsssl/scheme in xml/xsl it
starts to get really really ugly.

	Suppose Jade were given a new personality: an XML version with the
pattern/matching query language support i.e. DSSSL + XQL?

	Alternatively suppose the stuff inside <xsl:eval> where scheme? The problem
clearly is not the parens ... look at any xsl sheet and tell me it is easy
to read! rather the problem is the lack of focus on outputting HTML which is
what the world wants to do whether or not you agree. I suggest direct
XML/HTML output not requiring a back-end just as XSL can do without going
through FO.

Jonathan Borden
http://jabr.ne.mediaone.net

>
>
> The DSSSList reached its second anniversary yesterday.  The second
> anniversary for a mailing list isn't as remarkable as its first
> anniversary (which is my excuse for getting the date wrong), but I'll
> say a few things anyway.
>
> The DSSSList currently has 568 subscribers.  We high-scored in
> December with 578 subscribers, then dipped until February, but
> subscriptions are rising again.
>
> The DSSSList has had over 3,700 messages in two years.  This year
> didn't have as many messages as last year, and the list has been quiet
> of late, but I don't expect this quietness to last.
>
> The question of DSSSL's future occupies me from time to time, as it
> occupied the list recently, but I always conclude that there are
> people doing real work and solving real problems using DSSSL, and that
> they will continue to use DSSSL while it is the best tool for the job.
> And, until XSL implementations do "provide the formatting
> functionality of at least DSSSL", DSSSL will be the best tool for many
> jobs.
>
> That's not to say that DSSSL implementations provide the full
> formatting functionality of DSSSL.  It is heartening that there are
> more DSSSL implementations than previously, but, for the most part, we
> are using simple-page-sequence, with its rudimentary headers and
> footers, only because that's what our tools provide, and we would like
> the ability to produce more complex pages.  The DSSSL's future is not
> assured while the DSSSL tools don't fully implement DSSSL.
>
> The DSSSL's future is also not assured while DSSSL is seen as an
> arcane and impenetrable language.  To get on my favourite DSSSL
> soapbox, we need more and better DSSSL documentation so that people
> have something to help them get started with DSSSL.  We also need
> DSSSL editing software that will hide the gory details, and all those
> parentheses, from casual users so they don't need to begin to
> penetrate the language.  XSL may not have DSSSL's parentheses, but it,
> too, will need a friendly front-end before everybody can use it.
>
> I keep mentioning XSL because I'm interested in both, and because
> comparisons between the two are obvious.  The question for many people
> is undoubtedly whether they should adopt DSSSL or XSL.  I doubt that
> the answers are always clear cut, since for the moment DSSSL has the
> edge in print applications and, with Jade's SGML backend, is capable
> of producing HTML, which is the current favourite output of XSL
> engines.  XSL already has more implementations, but as yet they don't
> do as much.  I develop applications with both, and I expect that I
> will do so for some time to come.  Long live DSSSL!
>
> Regards,
>


 DSSSList info and archive:  http://www.mulberrytech.com/dsssl/dssslist