Re: Proposal: accessibility revision for the img element...

At 03:04 +0200 UTC, on 2007-06-25, Charles McCathieNevile wrote:


> 1. The alt attribute replaces the image if for some reason it is not
> rendered.

Is it not possible to upgrade <img> to a container, so that richer fallback
content becomes possible? It would remove the practcal length restrictions
on ALT, and would allow the textual alternative to contain markup.

> User agents must make its content available (whether by tooltip,
> context menu option, or otherwise) to users [...]

I understand what you mean and like it, but it may need to be spelled out
more explicitly that the intention is that when the image *is* available, UAs
must still make ALT available, but in a different way than when the image is
not available.

I propose this more explicit phrasing of your first paragraph:

   1. The alt attribute replaces the image with a textual alternative if for
   some reason the image is not rendered. When the image is rendered, user
   agents must make the alt attribute's content available (whether by tooltip,
   context menu option, or otherwise) to users. When the image is not rendered
   user agents must render the alt attribute's entire content in place of the

Note that this also replaces "fallback content" with "textual alternative".
Considering the confusion even here about what makes "proper ALT text" I
think "fallback content" will be misunderstood by many authors. "Textual
alternative" seems a better phrase.


> The relevant section from what says it is an Editor's draft of 23 June, at
>, has
> the following:
> [...]
> I would append to the second paragraph the following:
> "Where the image cannot be displayed, the user agent should present the
> text content of the alt attribute in its place. The entire content of the
> attribute should be available."

Note that while I completely agree with "entire content", it contradicts
saying that UAs may present ALT text in the chrome (tooltips, status bar,
whatever). The length of such widgets tends to be quite restricted.

The same problem exists with the title attribute. It might be a good idea if
the spec would define a maximum length (in characters, if possible) for both
attributes, so that authors can be sure that conforming UAs will present
these attributes' content in its entirety.

Sander Tekelenburg
The Web Repair Initiative: <>

Received on Monday, 25 June 2007 11:04:34 UTC