Re: @longdesc scope (was: HTML Media Transcript, Issue-194: Are we done?)

On Fri, 29 Jun 2012 03:50:11 +0200, Leif Halvard Silli
<xn--mlform-iua@målform.no> wrote:

> 'Janina Sajka', Thu, 28 Jun 2012 12:42:26 -0400:
>> John Foliot writes:
>>> Charles McCathieNevile wrote:
>>>>
>>>> I think this is a bad idea. It is stretching ourselves in
>>>> loops that are unnecessarily complex, which I think is more
>>>> likely to more likely to lead to confusion than to
>>>> improvements in the web.
>>>
>>> What Chaals said.
>
> OK. Leaving @transcript, entering @longdesc: The points I made are
> valid with regard to what a @longdesc attribute can point to.

Sure.

> So if one cannot change the name of the red rose, then one should at
> least specify that it isn't always red.
>
> E.g. the long descriptions[1] for the CSSquirrel cartoon[2] are often
> better described as transcripts than as long descriptions. In fact,
> there is a comics search engine (which searches a database of comics
> transcriptions) that talks about comics transcription.[3]
>
> Thus, ideally, the proposed longdesc spec text should give better hints
> about what a "long description" and "long text alternatives" are - that
> it can in fact be a transcript. [4]

Maybe. I think that is relatively editorial.

I also regard longdesc as a candidate for obsoleting eventually, if
aria-describedat or something actually gets take-up.

Given the general slowness of accessibility features being broadly
implemented (decades, where other HTML features are taken up widely in
maybe a year or so - although this matches the physical world where making
buildings accessible is done far slower than almost any other improvements
I can think of), I don't expect us to be in that position any time soon. A
lot of accessibility stuff unfortunately still seems to happen on a
different rhythm, more closely aligned with very large enterprises (who
generally seem far more likely to get around to implementing accessibility
as a normal part of their work) than the agile hackers at the bleeding
edge (of whom only a small minority seem to feel any need to provide it).
Trying to follow a group of people who aren't the main implementors is, as
we have acknowledged a zillion times in HTML5, something of a fools'
errand.

cheers

Chaals

Received on Tuesday, 3 July 2012 16:33:50 UTC