Re: feedback requested on WAI CG Consensus Resolutions on Text alternatives in HTML 5 document

Maciej Stachowiak wrote:
> On Aug 18, 2009, at 10:51 AM, Sam Ruby wrote:
>> Maciej Stachowiak wrote:
>>> As for whether we have an agreement in principle: I believe that we 
>>> do not have one yet. First, the PFWG reps on the call were very clear 
>>> that the ideas they described on host language semantics and ARIA 
>>> were tentative. I believe all agreed that PFWG should put their 
>>> thoughts on the record in email as the next step. Second, the PFWG 
>>> reps on the call agreed that the comments about the role attribute 
>>> were interesting, but did not state a position on whether ARIA would 
>>> be changed to reflect them, though they did agree to take these 
>>> comments back to PFWG. I believe once these two points are addressed, 
>>> then we will in fact have an agreement in principle. Does your 
>>> recollection differ from mine?
>> What I want to do is to break the logjam of "I won't respond until the 
>> other side puts their thoughts on the record".  I would like us to 
>> proceed based on the premise of an agreement in principle, even if it 
>> is subject to change.
>> The following is draft thoughts, subject to change, and we have the 
>> ability to influence these thoughts:
>> My recollection was that you thought this was sufficient, and upon 
>> reading Michael's post, I found nothing that surprised me.  But I am 
>> not as close to the problem as you are, let me know if you read this 
>> differently.
> I believe that email is sufficient to act on. I believe it is detailed 
> enough to have all the info needed to draft the HTML5 text for 
> integrating ARIA. I don't think it is reasonable to wait for an ARIA 
> Editor's Draft given this info. (That being said, it can't hurt to also 
> put it in the form of an Editor's Draft.)

Excellent! (and agreed).

>> More immediately, the status of issue-35 (aria-processing) is open. 
>> Cynthia Shelly had action 114 to get the PF working group to report 
>> progress.  Unless I hear differently, I am going to assume that that 
>> action is complete, and therefore want to know what the next action 
>> is, who owns that action, and when it is due.
>> Based on my (admittedly limited) understanding of the issue, a 
>> reasonable next step is for somebody to draft a matrix of potential 
>> combinations[1].  Even if some (or even many) of the squares in the 
>> matrix are wrong, the exercise will be useful in that that should 
>> spark useful discussion.
> If Ian doesn't want to do it, I suggest that Henri Sivonen should make 
> the first draft of such a matrix, since he's effectively already made 
> one for the HTML5+ARIA mode.

For reference:

If we have an agreement in principle, and a matrix, then the next step 
would be for somebody to produce a draft.  In any and all cases, we 
still need a task description, an owner, and a date.

It is clear to me that there will be second order issues uncovered by 
ARIA integration.  For example there appears to be a gap between what 
Steven and Maciej appeared to agree to (just yesterday!) and what Ian 
and Smylers believes (again yesterday!):

Note: I'm not suggesting in any way that this particular issue is a 
showstopper.  What I am saying is that ARIA integration is likely to 
uncover other issues, and finding out what they are is important.

Net: given that October is only six weeks away, I believe that 
accelerating progress towards an initial draft is in order.

> Regards,
> Maciej

- Sam Ruby

Received on Tuesday, 18 August 2009 18:26:14 UTC