Re: Backwards compatibility of new selectors (was: Color mo

Neil St.Laurent wrote:
> 
> > Not an argument for breaking bc.  As above - the spec is not easy to
> > follow.  It is not suprising that nobody has it right.
> 
> If you are referring to the grammar used by CSS then I could probably
> provide a real grammar for the language -- the one we are using to
> parse CSS2.  The grammar provided by the specification I find to be
> laid out very bad and has errors in it.

Two points on this:

1) That the grammer initially supplied was no good - it took alot
of work to make a yacc parser (for example) out of it.

2) That it is too late to offer to replace the grammer now that people
already have products based on the original grammer.  Perhaps this
is a good reason to have a sample implementation available of some
or all the stuff proposed by the group.  It is a strong motivator
and worked extremely well for the X Consortium.

Of course I don't have a problem with clarifications and such,
I'm mostly concerned that actions are taken with awareness of how
they affect real products, developers and users of these systems.
It is a mistake to develop standards in absence of this awareness.

Mind you,  I think that sometimes you do need to break things.  Not
this time though.

Doug
-- 
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 15:55:05 UTC