Re: The HTMLWG and WHATWG (Was: Default (informal) Style Sheet)

On Tue, 3 Apr 2007, Sebastian Schnitzenbaumer wrote:
> 
> The issue I've seen it that one group comes to consensus, passes it over 
> to the other group, the other group starts discussion again, comes to a 
> new consensus that may be different, passes it back to the first group, 
> the first group that already reached consensus re-starts discussion, 
> comes to a different consensus, passes it over...

The WHATWG doesn't use a consensus-based approach, so that wouldn't be a 
problem in this instance -- you'd only ever have one group trying to reach 
consensus. The WHATWG side would treat that group the same way it treats 
any other source of input.

To elaborate: in the WHATWG, each spec has an editor, who's responsibility 
is to take all feedback from all the sources I listed in an earlier e-mail 
(including blogs, forums, direct feedback from implementors, comments in 
bug systems, conclusions from relevant research, expert input, and, of 
course, discussions on the mailing list), and then put forward a proposal 
that addresses all the use cases, all the requirements, and all the 
constraints that have been put forward. (This has a side-effect of making 
it irrelevant if people agree or disagree with an individual's comments; 
only the quality of arguments, not the quantity, has any bearing on the 
result; thus, the WHATWG list has no +1/-1 posts).

The proposal is then reviewed by anyone interested in it, and further 
feedback is incorporated to address any issues found.

(There is also an oversight process: a small group of people, mostly 
browser vendors in the WHATWG case, have absolute authority to override or 
replace the editor should the editor start going off the rails. It hasn't 
yet been invoked, which I find encouraging.)

This process is the only way I know of to make significant forward 
progress in a group of hundreds of people, as far as I can tell (the 
WHATWG list has over 700 members). I would encourage the HTML group to 
adopt a similar approach, though I would be curious to see what other 
approaches would be like in a group this size.

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

Received on Tuesday, 3 April 2007 20:08:43 UTC