- From: Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>
- Date: Sat, 15 Jan 2011 03:06:40 +0100
- To: Benjamin Hawkes-Lewis <bhawkeslewis@googlemail.com>
- Cc: Doug Jones <doug_b_jones@me.com>, Maciej Stachowiak <mjs@apple.com>, HTML WG Public List <public-html@w3.org>
Benjamin Hawkes-Lewis, Sat, 15 Jan 2011 00:36:43 +0000: > On Fri, Jan 14, 2011 at 8:31 PM, Leif Halvard Silli wrote: >> SE[C]ONDLY: > I cannot follow your argument here. See my follow-up to Doug. [1] >> THIRDLY: No risks, you say. But perhaps there is a risk that authors, >> who could have increased their pages' accessibility by adding aria to >> their table based pages, just let their pages be as they are, because, >> after all, their pages causes no error in theor current state, while >> they would get validation errors if they added aria to their tables. > > False. [ snip ] True. [General:] I had not picked up that HTML5 had been updated to permit role="presentation" regardless of whether the element has strong [2] and weak [3] semantics. But I am not alone in not having picked it up: Validator.nu doesn't accept e.g. <html role="presentation"> yet Steve's CP hasn't picked it up either, it seems. [4] (Unless he simply disagrees with it.) At least, his CP explicitly lists the native semantics elements for which role="presentation" is permitted. [Table:] While it is true that HTML5 now permits <table role="presentation">, it isn't true for the reason you quoted [3] but because the <table> element was *removed* from [2] and [3] thanks to Ian's reaction to bug 10478 in late September. [5]. (I had not noticed that change either.) Thus, as a result, HTML5 ([2] and [3]) actually doesn't explicitly permit role="permission" for <table>. But this is also not necessary to do, since authors are free to use the ARIA attributes "in accordance with the requirements described in the ARIA specifications" as long as they do not conflict "with the strong native semantics described below" [2] or with the weak native semantics [3]. > (Conceivably, a hyper-intelligent checker might penalize authors for > abusing tables but it would penalize them regardless of the presence of > role="presentation", not because of it.) Good point. I had thought that Richard's point was about validation, primarily. But now I see that the entire point is an objection to HTML5 stating that tables should not be used as layout aid. >> As for the third, point, then I don't agree with myself. I tend to >> agree with Ian in that an honest @role is a 'godsend', which allows >> authors to check whether they have used an element for a valid purpose. > > What you mean like using an element with some sort of defined > semantic meaning, like a table for a table? Oh wait... ;) As I see it, the spec now more or less says what I suggested at the bottom of my follow-up to Doug: [1] ]] 1) it could saythat tables should not be used for layout - and nothing more. [[ [1] http://www.w3.org/mid/20110115003734617623.82ad2bb2@xn--mlform-iua.no [2] http://dev.w3.org/html5/spec/content-models#table-aria-weak [3] http://dev.w3.org/html5/spec/content-models#table-aria-strong [4] http://www.paciellogroup.com/blog/misc/HTML5/aria-html5-proposal.html [5] http://www.w3.org/Bugs/Public/show_bug.cgi?id=10478#c12 -- leif halvard silli
Received on Saturday, 15 January 2011 02:07:16 UTC