Re: ARIA Proposal

At 06:27 -0400 UTC, on 2007-10-01, Matthew Raymond wrote:

> Sander Tekelenburg wrote:
>> At 21:50 -0400 UTC, on 2007-09-27, Matthew Raymond wrote:


>>>    The |role| attribute is an elegant solution to a problem that doesn't
>>> exist. [...]
>> What about the problem Maciej raised in
>> <>, for
>> which @role would appear to be a solution? Does that problem not exist?
>    I'm not sure what you're referring to. There is no apparent use of
> |role| in the message you cite.

Forget it. It looks like I misunderstood what you were saying. (I merely
meant to say that the problem Maciej noted is real, and that @role might be
able to provide a solution.)

> Personally, I think a combination of
> <link> elements and |id| attributes would serve just fine if you want
> auto-assigned or user-assigned keys

Your example, <link rel="rolename" href="#myid"> [...] <div id="myid></div>,
looks interesting, yes. I have some doubts about it requiring authors to
specifiy their intention in two places though. <div meaning="value"> would be
easier to author, no?

Your example "<div namespace-rolename></div>" doesn't have that problem. But
it seems to me that it would be much easier for authors to have a a spec that
provides a single attribute with a list of possible values. Perhaps this
doesn't have to be a problem if the spec would list all namespace-rolenames
clearly together in one section though.


>    Yes, the |role| attribute could be used, but the point is not whether
> you can invent uses for |role|.

Agreed. That wasn't my take in
<>. Quite the
contrary. To use your terminology, Maciej's proposal "invented a use" for
@accesskey. My response was essentially that we should look better at the
actual problem and that *something* that allows authors to markup the meaning
of content would probably allow for much better key-combo's in UAs (because
not author-defined). I don't care whether the solution is @role or something
else -- just that it is a good solution.


Sander Tekelenburg
The Web Repair Initiative: <>

Received on Friday, 5 October 2007 01:25:27 UTC