W3C home > Mailing lists > Public > www-style@w3.org > July 2005

Re: The Progress of CSS

From: Ian Hickson <ian@hixie.ch>
Date: Sat, 2 Jul 2005 01:34:36 +0000 (UTC)
To: Adam Kuehn <akuehn@nc.rr.com>
Cc: www-style@w3.org
Message-ID: <Pine.LNX.4.61.0507020127500.13594@dhalsim.dreamhost.com>

On Fri, 1 Jul 2005, Adam Kuehn wrote:
> 
> Well, for example, does a draft really need a full six months in CR?

The idea of the six month clause is that if we would have met the other 
conditions before the end of six month, we would wait six months to make 
sure no issues came up.

We (the CSS working group) have never had a spec meet its CR exit criteria 
within six months of publication.

So removing the six month clause would have no effect.


> If you really think the answer is yes, perhaps it could receive ONLY six 
> months.  The current version of CSS2.1 is back to WD status after 
> spending SEVENTEEN MONTHS in CR.

You seem to think that the idea of CR is to just have the spec sit in CR 
then move to PR. This is not the case. The idea is to sit in CR *until 
there are two complete interoperable implementations of every feature*.

Just proving this requires a detailed test suite, which will itself take 
years to produce.


> The original WD spent a year and half in WD status, albeit in three 
> different incarnations.  It seems to me that each step needs to be 
> shorter.

The time that the spec spent in WD status was the shortest possible time 
we could make it while still actually resolving all 900 issues that were 
raised over that time period.


> Does a CR need two COMPLETE, interoperable implementations to become a Rec?

Yes, how else could we prove that the specification is interoperably 
implementable?


> There aren't really more than four significant vendors in the field.  
> Why require 100% compliance from half the known universe?

I'd _like_ to require 100% compliance from the entire known universe but 
for the sake of proving that it is possible to implement the spec 
interoperably, we only need two.


> Finally, if the W3C is going to publish a schedule, as it has for the 
> various CSS3 modules, it ought to try to at least come close.  The only 
> modules which have the Test phase even scheduled were all supposed to 
> hit Rec status in '04. None have.  Indeed, halfway through '05 we have 
> zero Recs, despite the fact that no fewer than 20 items in that table 
> are scheduled for Rec status by the end of this year.
> 
> This is not a good track record.

Yeah, our published schedule is out of date. There is a new one which is 
somewhat more realistic (although still, in my opinion, rather optimistic) 
but it is currently member-only for some reason.

The CSS working group is chronically undermanned. If you know of a W3C 
member organisation that can contribute competent editors to the CSSWG, 
I'm sure they'd be welcome. We certainly have plenty of work for editors 
to do.

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'
Received on Saturday, 2 July 2005 01:34:40 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:39 GMT