Re: [link-relations] NEW RELATION REQUEST

[Ian Hickson <[e-mail]>]

On Wed, 11 Aug 2010, Julian Reschke wrote:
> > > 
> > > That being said, I'm fine with this registration if it actually 
> > > refers to the W3C Working Draft (that 
> > > is,<http://www.w3.org/TR/html5/>).
> > 
> > So to be clear, you are rejecting this registration unless it 
> > references the out-of-date snapshot document on the W3C site rather 
> > than either the WHATWG document or the current W3C document? (So in 
> > particular, you are rejecting this registration at least until the 
> > next snapshot publication?) ...
> 
> I haven't rejected anything. I'm saying that the registration should 
> point to a stable document in W3C space, and I don't think the "current 
> editor's copy" qualifies as that.
> 
> What's the problem with waiting with the actual registration until the 
> new WD is out?

The TR/ page drafts are out-of-date obsolete drafts. We shouldn't be using 
them as a reference, as they are almost by definition known to be 
incorrect. (I don't believe we have ever published a TR/ draft of this 
specification that was up-to-date when it was published.) I would 
definitely not consider it acceptable to have the registration point to a 
document that is known to be wrong.

It's not clear what benefit it would bring, either. That draft is no more 
stable (it changes frequently), it's no more representative of consensus 
(we have always had more outstanding known issues when publishing than 
the contemporary editor's draft), and it is technically inferior (again, 
it has always been published with more known issues).

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