RE: ARIA BPG Issue: we should not be decreeing complex keyboard interactions

Well, I do not speak of non-accessibility of grids, I speak of different
implementations for keyboard navigation in different grids based on MS
technology.

And yes, I believe ANY written basic reference standard for speech
output is better than proprietary approaches or having nothing like
that. Sad, bud we disagree here. 

The problem is not a central definition.  The problem is that these
definitions by standards bodies tend to have the technical charme of a
dustbin design and  come toooo slooooowwww for industry's "We need it
NOW and fancy" approach ... 

Isn't it always better to watch "Tool Time"-like series and
Do-It-Yourself instead of waiting all month long for the craftsmen ? :)
.. and don't forget: too many cooks spoil the broth ..

- Stefan

-----Original Message-----
From: James Craig [mailto:jcraig@apple.com] 
Sent: Samstag, 6. September 2008 01:15
To: Schnabel, Stefan
Cc: wai-xtech@w3.org
Subject: Re: ARIA BPG Issue: we should not be decreeing complex keyboard
interactions

Stefan Schnabel wrote:

> Both Agreed. But what will you do if there is no keyboard standard  
> even
> for the same control type on desktops?
> Imagine 20 3d party WPF-based MS grid controls that work differently
> here .. what to take as "standard"?

Perhaps I don't understand your question. It's Microsoft's call as to  
how to standardize their controls like a WPF grid. I thought the point  
of WPF was that the control could look like anything, but the  
programming framework would understand its semantics. If that's the  
case, then the framework should hand those semantics off to MSAA. If a  
grid isn't accessible in a desktop app, how would its equivalent be  
made accessible on the web?

> I do not think that a proprietary AT navigation for aria roles is a
> blessing for the web.
> The problem is that in the past this definition part has been left
> totally to the AT vendors and you know the outcome.

I'd disagree with you here. I am very thankful that VoiceOver's  
interaction model was not designed by a standards body.

James

Received on Monday, 8 September 2008 07:43:28 UTC