Re: request to the HTML WG Chairs on ISSUE-129 aria-mapping

On Mon, 11 Apr 2011, Maciej Stachowiak wrote:
> On Apr 11, 2011, at 2:01 PM, Ian Hickson wrote:
> > On Mon, 11 Apr 2011, Maciej Stachowiak wrote:
> >>
> >> This doesn't mention a default role, unlike the text for "h1 to h6 
> >> element that does have an hgroup ancestor". One could plausibly 
> >> interpret this as:
> >> 
> >> (a) Change h1-h6 to have no default role
> >> (b) Make no changes to default role for h1-h6
> >> 
> >> From the test alone, it is not completely clear which was intended. 
> >> From your statements, it seems that your intent was (b). And it seems 
> >> that Ian agrees that (b) would have been a technically superior 
> >> choice.
> > 
> > (b) would be self-contradictory (making the default non-conforming).
> 
> 1) There doesn't seem to be a technical reason that the implied default 
> value must also be allowed as an explicit value. So I don't see the 
> contradiction.

Wouldn't it mean that every document with a heading was implicitly 
non-conforming?

I certainly wouldn't describe it as "superior", anyway.

I agree that my assumption here (that the removal of 'heading' from the 
list of allowed values also meant that it couldn't be the default) was 
unwarranted, though.

Anyway, I'll revert the change as requested.

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

Received on Monday, 11 April 2011 23:19:56 UTC