- From: Jason White <jason@jasonjgw.net>
- Date: Wed, 8 Aug 2007 20:01:20 +1000
- To: public-html@w3.org
On Wed, Aug 08, 2007 at 07:06:53PM +1000, Ben Boyle wrote: > I would like the editors to consult with WCAG when they review this > part of the specification. > > I'd personally prefer meta@refresh be deprecated. In my work I'm bound > by WCAG so it essentially is deprecated (for me) anyway (and maybe > many of you feel that is sufficient - it isn't, imo). I'd prefer HTML > 5 be aligned with WCAG (it almost always is). In lieu of this, a > simple acknowledgement would go a long way: how about a Note > referencing WCAG checkpoints 7.4 and 7.5. (I still think a more > well-rounded outcome will come from consulting the WCAG group. There's a strategic choice to be made here. Option 1: specify refresh in HTML 5, in a way that takes into account considerations deriving from accessibility requirements. At a minimum, it must be possible to conform to both HTML 5 and the User Agent Accessibility Guidelines 1.0, in which the notable requirement here is checkpint 3.5: http://www.w3.org/TR/uaag10/guidelines.html#tech-configure-content-retrieval Preferably, HTML 5 should encourage implementation choices that are consistent with UAAG 1.0, not only in the present context but also in connection with other algorithms and statements in the spec that affect the user interface and relevant APIs. Option 2: deprecate refresh. The risk here is that scripts would then be used by authors to achieve the same result, reproducing exactly the problems that have been raised in this thread. Also, since HTML has historically allowed user agents to interpret documents which include implementation-specific extensions, and this permissive approach is likely to continue in HTML 5, the consequence of deprecating refresh is likely to be that user agents would implement it anyway, albeit as an extension, possibly without taking the difficulties discussed here into consideration. It might be better, therefore, to specify refresh, but to make it compatible with UAAG 1.0. As it stands, I don't think step 22 of the algorithm stated in the HTML 5 draft is compatible with UAAG 1.0 or with accessibility-related concerns.
Received on Wednesday, 8 August 2007 10:01:39 UTC