W3C home > Mailing lists > Public > public-xhtml2@w3.org > October 2007

Re: direct link to latest version of S. Pieters' ARIA Proposal

From: Anthony Grasso <anthony.grasso@cisra.canon.com.au>
Date: Sun, 07 Oct 2007 00:48:27 +1000
Message-ID: <4707A03B.1050805@cisra.canon.com.au>
To: Doug Schepers <schepers@w3.org>
CC: public-xhtml2@w3.org, Simon Pieters <simonp@opera.com>, aleventh@us.ibm.com, www-svg <www-svg@w3.org>

Hi Doug,

Doug Schepers wrote:
>>
>> So as far as I can tell 'roles' are specified on elements and it kind 
>> of categorises them? right?
> 
> Right, the 'role' attribute can be used to simply provide some semantics 
> without any behavior, so you could label something as related to 
> navigation, or as a banner ad, or pretty much any other categorization 
> that people find useful.  The 'class' attribute could also be used this 
> way, but that's more controversial, since some people think that CSS 
> should only be used for styling, and that attaching semantics to it adds 
> content.
> 
> 
>> If the above is correct, then I guess this gives implementations the 
>> option to treat elements in a certain way according to their 'role'?
> 
> Right again, in combination with the ARIA taxonomy, a UA can assign 
> particular preset behaviors (including supplying user notifications and 
> choices of actions) to elements so labeled.
> 

Ahhh ok. I see. That makes sense.

> 
> 
>> Or again if the above is correct, a use case would be having an XSLT 
>> to remove all of the elements that have 'navigation' 'role' when 
>> printing?
> 
> I don't think that's a primary use case in the minds of the designers, 
> but neither do I think it's a violation of the spirit of it.  In fact, 
> the 'role' attribute could take on special values that pertain 
> specifically to print, and don't affect the on-screen rendering or 
> interactive behavior at all.
> 

Ahhh, of course.

> As I understand it, 'role' is meant to be pretty generic, though the 
> XHTML2 WG has some clever uses for it already in mind.
> 

Ok, nice! I'm curious to see the use cases that XHTML2 WG have - mainly 
because this sounds very interesting.

> 
>> Is my understanding correct... or have missed the boat completely?
> 
> No, I think you just caught a different boat, but it's also traveling in 
> a useful direction.  It's good to have people thinking in different 
> directions.  That 'role' can be useful to offscreen uses (not just 
> print, but pre- or post-processing, and for a variety of tasks with 
> XSLT) reinforces that it is is useful to add.
>

No worries. It does seem like that it's something handy to have around. 
Thanks for the feedback. I have a better understanding for the purpose 
of 'role'.

Cheers,

Anthony
Received on Saturday, 6 October 2007 14:48:55 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 23 February 2010 18:12:46 GMT