W3C home > Mailing lists > Public > www-style@w3.org > December 1997

Color models and CSS2 in general

From: Douglas Rand <drand@sgi.com>
Date: Tue, 02 Dec 1997 10:11:28 -0500
Message-ID: <34842520.9796CDDB@sgi.com>
To: www-style@w3.org
What I want to know is why this is being reinvented in this 
venue?  Why not look at the substantial work done in,  for example,
X for device independent color specification (CMS) and use that in
CSS2?  There isn't a good reason *not* to include all the relevant
models including RGB, HSL, CIE*, etc.  They can all be more or less
transformed to each other or a common internal format and we aren't
talking about anything significant in development time or code space.

On a slightly different topic which is related in my mind,  I'm
concerned with a few things about css2:

1) Adding specifiers which are dependent on parents and siblings is
making a non-backward compatible change to the spec.  I don't remember
a hook which made things with funky syntax be rejected.  Users with
first generation agents are going to be pretty upset when CSS2 pages
render poorly on their browser.  In fact the syntax for CSS2 had better
be forward and backward compatible.  If we must have positional syntax,
make it look like an existing specifier:

P B:child[P] { ... }
P B:child[P,first] { ... }
P:sibling[P,previous] { ... }

At least this will be ignored by CSS1 implementations (in the sense
that these will never match). 

2) Problems which have long been solved in DSSSL (and now XSL) are being
resolved in this arena.  Especially as CSS tries harder and harder to
deal with print media,  I think the holes in the model will
become progressively more apparent. Perhaps it would be wiser, not
that I expect anyone here to listen to me, to demure on these higher 
end features in CSS? Making sure XSL can subsume CSS1 syntax might
be more productive.  Aren't both Netscape and Microsoft intending
to support XML fully?

3) I strongly object to the development process for the CSS(n)
specification in general.  I very much dislike having to raise this
sort of objection at such a late date in the process.  A more open
process worked very well for Motif and the X toolkit intrinsics, 
similarly such a process worked well on the single ANSI committee I took
part in.  I do not see why that cannot work here.

Doug Rand				drand@sgi.com
Silicon Graphics/SSO			http://reality.sgi.com/drand
Disclaimer: These are my views,  SGI's views are in 3D
Received on Tuesday, 2 December 1997 10:15:12 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:26:45 UTC