- From: Adrian Roselli <Roselli@algonquinstudios.com>
- Date: Tue, 4 Sep 2012 20:14:46 +0000
- To: Mathew Marquis <mat@matmarquis.com>, Leif Halvard Silli <xn--mlform-iua@xn--mlform-iua.no>
- CC: Peter Winnberg <peter.winnberg@gmail.com>, Steve Faulkner <faulkner.steve@gmail.com>, HTML WG <public-html@w3.org>, "public-respimg@w3.org" <public-respimg@w3.org>
> From: Mathew Marquis [mailto:mat@matmarquis.com] > > Based on the feedback we've received in both this thread and a > quick informal poll of developer preference in the respimg CG's > mailing list ( archived here: > http://lists.w3.org/Archives/Public/public- > respimg/2012Aug/0045.html ), it seems that while it's generally > accepted that the `aria-describedby` approach is the more elegant > of the two approaches, the duplicated `alt` attribute is the more > intuitive. It also stands to have the `alt` information available > to the widest number of users, where `aria-describedby` may not be > available in all contexts. It's unfortunate, but our obligation is > to the end users first and foremost. > > I think our best bet is to require that authors specify an `alt` > attribute on both `picture` and the fallback `img`. There has been > a great deal of discussion along these lines to support that > decision, as outlined in Laura's incredibly helpful message here: > http://lists.w3.org/Archives/Public/public- > respimg/2012Aug/0047.html I don't take Laura's message to say that @alt should appear on both <picture> and <img>, just that the ARIA role isn't a fit. I only mention this because I didn't read her message as an endorsement of a double-@alt, just on @alt over ARIA. In your survey to the community group you outlined two options. I know I saw at least one other person (who wasn't me) propose a third. Which you start to touch on next... > I am still curious about the viability of allowing an `alt` > specified on the fallback `img` to "bubble up" to the parent > `picture` element - if that should be possible, I think it would be > ideal if we could mention that specifying an `alt` on the fallback > `img` applicable to the parent `picture` is also a valid approach. I still see @alt on *both* <picture> and <img> to be at risk from copy-paste errors (at least until authoring/WYSIWYG tools catch up). Even if not that, when someone changes the value of one it seems likely to me that the other could (will) be missed. I'd rather see <picture>'s fallback rely on the existing momentum <img> has with its @alt -- just rely on <img> to be the fallback both for the alternate image and the @alt text. Leave @alt off <picture> altogether. I am also trying to look at this in a vacuum, without bringing <figure> into play and without drawing comparisons to <object> and <canvas>, partly because so many young web devs I know have no concept of how those elements work and aren't in a position to make the same analogous connections we are. Or am I missing something fundamental here?
Received on Tuesday, 4 September 2012 20:15:14 UTC