Re: SGML/XML syntax for DSSSL

[Paul Prescod <papresco@xxxxxxxxxxxxxxxxxxxxxxxxx>]

James Clark wrote:
> 2. An alternative infix syntax for the entire language; Dylan seems to me
> the obvious starting point.

Looking at Dylan's syntax for creating new objects I would guess that
James has *already* been looking at it. =)

This is Dylan code to make a new object:

 make(<actor>, name: "Betty", fans: 5000000);

Dylan looks pretty good.

http://legend.gwydion.cs.cmu.edu/dylan/interim/interim-18.html

As Lee pointed out the Dylan "select" statement syntax is closer to Perl
than the one I described today (but still quite elegant). The "let"
syntax is similar to the one I used.
 
> I don't see these alternatives as mutually exclusive.  They're addressing
> different audiences: (1) is addressing people who aren't programmers; (2) is
> addressing people who are. I'm not (yet) saying I think these are both good
> ideas, but I think they are both technically possible.

I'm not yet saying that either. =) I am sensitive to Earl's worry about
bifurcating languages once we hand control over to the W3C. 

A big question in my mind is if we hand Netscape and Microsoft the
existing DSSSL syntax will they implement it? Netscape in particular has
shown an incredible level of stupidity in the past. A java-like syntax
would serve as some sugar-coating (please pardon the pun). James may be
in a better position to investigate whether regular DSSSL will "fly"
with the W3C and vendors. If not, I would rather real DSSSL competed
with some kind of Dylan-DSSSL than with "Javascript stylesheets" which
are based on totally different concepts and models. Even if ylan-DSSSL
gets bastardized in some future version we are further ahead than
competing with "Javascript."

 Paul Prescod

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