- From: Charles Pritchard <chuck@jumis.com>
- Date: Mon, 13 Feb 2012 00:34:50 -0800
- To: Benjamin Hawkes-Lewis <bhawkeslewis@googlemail.com>
- Cc: Jonas Sicking <jonas@sicking.cc>, Maciej Stachowiak <mjs@apple.com>, Laura Carlson <laura.lee.carlson@gmail.com>, Sam Ruby <rubys@intertwingly.net>, "public-html@w3.org" <public-html@w3.org>
On Feb 13, 2012, at 12:02 AM, Benjamin Hawkes-Lewis <bhawkeslewis@googlemail.com> wrote: > On Mon, Feb 13, 2012 at 5:24 AM, Charles Pritchard <chuck@jumis.com> wrote: >> I think we've reached an agreement regarding your proposal: @hidden currently follows display: none, from CSSOM and your proposal seeks to have @hidden have its own properties. > > Jonas's proposal brings the conformance requirements in HTML in line > with ARIA's requirements and implementation guides. Please review: > > http://lists.w3.org/Archives/Public/public-html/2012Feb/0010.html > > I don't think we should ship HTML5 in direct contradiction of ARIA > without (a) trying to get WAI to engage on these issues or, failing > that, (b) explicitly marking wilful violations. So I certainly think > this @hidden issue needs to be resolved before Last Call, not deferred > to ARIA 1.1 and HTML6. > > Authors and implementors have to grapple with these contradictions > irrespective of the W3C's gradual snapshotting of the Living Standard. WAI-ARIA is an independent/agnostic language. That it specifies an agent walk a hierarchy is part of its design. There have been many attempts to reconcile ARIA and HTML5 through limiting ARIA expressions. These often include limiting the application of role on particular HTML5 objects. I find these attempts to be a misunderstanding of boundaries. The tension-- not contradiction -- is about interactive elements. There seems to be some thought that @hidden elements ought to be interactive (for the UA). I disagree. Regardless-- ARIA follows whatever the host has in place. And at this point, HTML5 @hidden rests on CSS display none. I'm happy with that shortcut, though I do sympathize with the impulse to mix visibility: hidden and display: none semantics. I'm just quite concerned about it, as it as been an effective boundary. That stated, I have every intention of using <canvas> to create accessible paths, both visually and semantically rich. I believe that HTML will not have another feature-rich alternative until further work and integration happens with SVG2. I can cheat, so to speak, with Canvas, and with longdesc. My needs are met. I'm disclosing that as a point of bias. I do not believe my bias is distorting my understanding of ARIA. -Charles
Received on Monday, 13 February 2012 08:35:19 UTC