- From: Jason White <jason@jasonjgw.net>
- Date: Fri, 9 May 2014 18:37:15 +1000
- To: wai-xtech@w3.org, "w3c-wai-pf@w3.org WAI-PFWG" <w3c-wai-pf@w3.org>
James Craig <jcraig@apple.com> wrote: > 4. <input type="checkbox" aria-required="true"> [...] > > Example #4 is invalid and not required because there *is* an html:@required > attribute available on html:input, so the lack of this boolean attribute > means the element is not required. The explicit ARIA attribute > aria-required="true" is in direct conflict with this value, so the host > language attribute should win. See HTML bug #23376. > https://www.w3.org/Bugs/Public/show_bug.cgi?id=23376 This is true. The strong native semantics table in the HTML 5 spec effectively prohibits example #4, if I am reading correctly. If I understand rightly, however, one could write example #4 in an HTML 4.01 document, where @required is not supported. Thus those who want example #4 to be allowed for backward-compatibility can have it, as long as they're not using HTML 5. This seems entirely reasonable to me - if you're writing in HTML 5, then you're assuming an HTML 5-conforming user agent, which must support the required attribute on input elements. The rationale is to make it impossible for a conforming document to be in a state whereby: <input type="checkbox" aria-required="false" required> and this is done by specifying strong native semantics and thus disallowing the aria-required attribute in such cases.
Received on Friday, 9 May 2014 08:37:44 UTC