Re: [webcomponents] Custom Elements Spec

On Mon, May 7, 2012 at 8:09 PM, Anne van Kesteren <annevk@annevk.nl> wrote:
> On Mon, May 7, 2012 at 9:04 PM, Ian Hickson <ian@hixie.ch> wrote:
>> If there's better ways we should certainly consider them. We have to bear
>> in mind though that fallback is something most authors don't spend much
>> time on, so the ideal is that having good fallback is _easier_ than not
>> having fallback. We should definitely not end up in a state where you have
>> to do _extra_ work to get fallback. (This is one of the big problems with
>> the way the ARIA spec is designed; it assumes that authors are going to
>> first spend loads of effort making a widget out of <div>s, and _then_
>> they're going to do even _more_ effort to annotate it with ARIA roles. We
>> don't want to end up in that kind of situation.)
>
> Yeah, I am similarly pessimistic that relying on ARIA to make
> components accessible is not going to work. That is, it could work if
> people used it, but given the complexity of doing so the common case
> will be inaccessible components. And that's bad.

Yes! Inaccessible components are bad. But I am optimistic that what
Rafael is advocating is slightly better than the alternative. FWIW, I
Iooked at it a year or so ago
(http://wiki.whatwg.org/wiki/Custom_Tags_Analysis#Accessibility), and
distinctly remember the notion of "killing bad patterns by baking
accessibility into components" as a hopeful sight of light at the end
of the currently available tunnel.

If you look at the two alternatives, one (the "is" attribute) asks the
authors to make the right choice. The other asks the component
developers to make the right choice. In the former case, the pool of
people who need to do the right thing is several orders of magnitude
larger than the latter. From there, it takes pure statistics to figure
out which alternative is likely to get better results.

:DG<

>
>
> --
> Anne van Kesteren
> http://annevankesteren.nl/

Received on Tuesday, 8 May 2012 03:49:05 UTC