- From: Maciej Stachowiak <mjs@apple.com>
- Date: Tue, 16 Mar 2010 02:31:37 -0700
- To: Steven Faulkner <faulkner.steve@gmail.com>
- Cc: Richard Schwerdtfeger <schwer@us.ibm.com>, Michael Smith <mike@w3.org>, public-html-a11y@w3.org
On Mar 16, 2010, at 2:17 AM, Steven Faulkner wrote: > Hi maciej, > > >To reframe my earlier line of questioning - is it conforming to > specify navsubtree=false? > > yes, for example if there is a canvas based chart along with a html > table of the chart data: > > <canvas navsubtree="false"> your browser does not support the > graphical rendering of the data, browser x or y is required</canvas> > > <table> > chart data > </table> > as in the current spec, the accessible alternative does not have to > be provided within the canvas. in the example above the accessible > alternative is provided outside of the canvas, the canvas content is > fallback. That seems like a reasonable design. My only suggestion would be to make the attribute a more typical HTML boolean attribute; for most of them, only presence or absence matters, not the value. Since navigable subtree is the default, perhaps the attribute should be "nonavsubtree" or "nonav" or something like that, and then it can just be a normal HTML5 boolean attribute. (Those names don't read so great, but I think mixing the false/true pattern with the present/absent pattern is confusing). Regards, Maciej
Received on Tuesday, 16 March 2010 09:32:13 UTC