Re: Language choice (was: Re: Interactive XML)

Subject: Re: Language choice (was: Re: Interactive XML)
From: Brandon Ibach <bibach@xxxxxxxxxxxxxx>
Date: Mon, 29 Jun 1998 19:47:26 -0500 (CDT)
Gregg Reynolds said:
> Brandon Ibach wrote:
> > 
> > So, could someone
> > briefly explain, or point me to something online which explains, why
> > a whole new standard is being created rather than some type of
> > adaptation or profile of DSSSL?  TIA...
> > 
> I've posted similar anguished pleas and the usual response is "it was
> thought that..." or "it was decided that..." DSSSL was too hard, or too
> "printy", or something along those lines.  Certainly DSSSL can be
> improved, and it doesn't even address large areas of functionality of
> great importance to the online world.
   As for the "too hard" business, see below.  As for DSSSL's
shortcomings for online use, I'll grant some problems.  However, it
seems like the areas where the problems exist are the very ones we've
just seen a huge discussion of: interactivity.  The consensus seems to
be that interactivity is important, but the XSL team has other things
to worry about as well, like hung punctuation.  Aren't things like
that the very things that DSSSL has already addressed?  Wouldn't time
be better spent addressing shortcomings in DSSSL, producing a W3C
recommendation for extensions/modifications, then giving it to ISO to
worry about while we all get on with implementing this stuff?

> But I've never seen or even heard
> of any reasoned critique of it (ok, I may not have been paying
> attention, but if it's out there it's pretty well hidden).  I'd love to
> see an account of exactly where "it is thought" that DSSSL falls short,
> but I think we'll just have to infer this from XSL publications.
   Ditto... some hard facts to back up abandoning DSSSL would be nice.

> There
> is some merit to the "too complicated" charge, but there is also some
> merit to the argument that it, like many of its ISO cousins, is rendered
> far more complicated than necessary by the language in which it is
> written.  And I don't mean English.  The other complaint I've seen is
> the old "too many parentheses" whine, er (sound of hyperventilating),
> argument against Scheme.  
   Hey, I like Scheme, but I'll also admit that my eyes hurt if I look
at it for too long.  It's not the friendliest language out there.
But, it's damn well suited to the task.  So, people want an HTML-ish
looking style language?  Standardize a mapping of HTML-ish syntax to
standard DSSSL!
   Let's face it... perception counts for a lot.  Ever introduce a
newbie to HTML?  "Ugh... that looks really complicated!"  Well, it
looks that way at first glance, but once you get into it a little,
it's really not bad at all.  Some, however, won't go there, so we
created HTML editors.  If you don't want to do DSSSL, then use a
simpler syntax that maps to it.  When you're ready to move on, or when
you need more functionality, then dive in.

> And remember that DSSSL is an ISO standard.  Great, international
> consensus, but imagine how long it would take ISO to do an upgrade of
> DSSSL.  Much better to have a strongly managed, quick and productive
> organization like W3C do it; just pretend that ISO commissioned W3C to
> adapt DSSSL to the web.  And in truth I think once the whole thing is
> finished you'll find that indeed XSL is "some type of adaptation or
> profile of DSSSL".  
   Granted.  ISO does its job well, but slowly.  So, as mentioned
above, let the W3C hash it out quickly so we have a standard, of
sorts, then give it to ISO as a wonderful head start.

-Brandon :)

 XSL-List info and archive:

Current Thread