XSL: Strawman comments, XS-970522

Subject: XSL: Strawman comments, XS-970522
From: Chris Maden <crism@xxxxxxx>
Date: Thu, 21 Aug 1997 11:32:21 -0400
3.1 aligned-column

Do we really need this feature?  To me, it seems unnecessarily fancy;
can a supporter please justify it?

3.3 Marginalia

Again, this seems unnecessary for a minimal conforming application.
Can someone state why it's needed?

4.5 font-name

The font-name characteristic takes a public identifier as its
argument; the strawman dictates that it shall be handled in the manner
mandated by the Standard.  This means that XSL will be caught up in
defining a method of resolving public identifiers to some object, and
specifying how those objects shall describe a set of font
characteristics.  In general, the emphasis of DSSSL on public
identifiers, while being the SGML way, seems to conflict with the
URL-centric XML way.

4.6 first-line-align

See 4.15.

4.9 input-whitespace-treatment

Is my thinking correct that a requirement of support for this
characteristic obviates the need for XML-SPACE in XML?  If so, can
someone more expert in DSSSL than I vouch for this in the XML SIG?

4.12 notation-system-id

The strawman requires this to be handled in the standard; the standard
mandates that this must be specified, and must be a system
identifier.  I repeat the plea I have made on the XML SIG; notation
system identifiers must be permitted to be MIME types.  This
requirement will frustrate creators of XSL stylesheets; what do I put
for the system identifier for GIF?  People will make up bogus things
and applications will ignore them, unless we permit a reasonable value
here.  If you agree, please speak up on the XML SIG; I feel that this
point is being ignored.

4.15

In 4.6, it was said that first-line-align can be ignored.  That is not
possible if the alignment-point flow object class must be supported.
I still think that aligned columns are not necessary in minimal
implementations and would favor removing this clause, but if it is not
removed, 4.6 needs to be changed.

4.17 table-auto-width-method

As with 4.5's font-name, I feel that this will drag XSL into needing
to define public identifier mechanisms.  For this particular
characteristic, it would be possible to enumerate a list of public
identifiers and methods (HTML method, CALS method), but it must be
handled carefully and not, I think, with S as in the strawman.

4.23 background-tile

This requires a public identifier for a graphic, at odds with XML's
URL-centric view, and with established Web practice of referring to a
background graphic with a URL.  This will seriously upset experienced
Web authors, and will again drag us into specifying a public ID
resolution method.

-=-=-=-=-

-Chris
-- 
<!NOTATION SGML.Geek PUBLIC "-//Anonymous//NOTATION SGML Geek//EN">
<!ENTITY crism PUBLIC "-//O'Reilly//NONSGML Christopher R. Maden//EN"
"<URL>http://www.oreilly.com/ <TEL>+1.617.499.7487 <FAX>+1.617.661.1116
<USMAIL>90 Sherman Street, Cambridge, MA 02140 USA" NDATA SGML.Geek>


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


Current Thread