Re: How Do I Propose a CSS3 Spec Change?

On Thu, 10 Nov 2005, Lachlan Hunt wrote:
> Hypothetically speaking, say a really good specification of an original 
> idea for a new property has been drafted and sent to this list.  The 
> spec clearly describes in great detail about its use cases, how it 
> affects rendering, how it interacts with other properties, the box model 
> and/or other relevant sections of the CSS, plus describes how it can be 
> incrementally rendered, any parsing issues, cascading and inheritence 
> issues, etc.  The author also satisfactorily responds to all questions 
> and comments raised.
> What are the chances of such a proposal at least being discussed for 5 
> minutes (before being rejected) at the next CSSWG meeting?

At the _next_ meeting, almost zero.

However, if it is a suggestion for a spec that I am an editor of (at the 
moment that would be CSS2.1, Selectors, Lists, and Generated Content), it 
is almost guarenteed to be discussed at a future meeting. Issues raised 
and proposals made get added to a first-in-first-out issues list, and 
issues remain on the list until they are resolved. Even minor issues 
frequently get significantly more than five minutes of discussion time 
(many relatively minor issues have had entire man-hours of time dedicated 
to them). Only issues that can be completely resolved by referring to 
previous working group consensus would be dismissed without the working 
group discussing the issue (in such a case, the issue would be responded 
to on the mailing list).

In particular, every CSS2.1 issue raised on this list, as well as those 
raised indirectly through comments in the Mozilla, Webkit, and Opera bug 
systems, is guarenteed to be looked at. This includes comments that 
weren't intentionally raised as CSS2.1 issues but which unintentionally 
indicated an error, inconsistency, or ambiguity in the spec. (Since I 
started recording CSS2.1 feedback we have resolved around 1200 issues.)

Similarly, all the Selectors ideas, even those that were dismissed as 
crackpot ideas on the mailing list, have been looked at by at least one of 
the Selectors editors. The good ones have been further discussed at face 
to face meetings. Proposals such as those on how to deal with table 
columns in selectors have been given extensive thought. (In this 
particular case, though, don't expect anything soon -- Selectors 2 is 
pretty much the last priority of the working group, and there is about a 
decade's worth of other CSS work scheduled before it.)

I can't speak about specs I'm not an editor of. In the CSS working group 
it is mostly the responsibility of the editor(s) of a spec to decide how 
feedback for that spec is handled. Editors can change, too, especially for 
some of the specs that we don't consider as immediately important, and 
when an editor changes past feedback usually slips through the cracks, 
even assuming it was being collected at all.

Ian Hickson               U+1047E                )\._.,--....,'``.    fL       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'

Received on Thursday, 10 November 2005 06:52:01 UTC