This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
Examples and bullet they seem so support: - img with alt text (E) - img with alt text and labelledby (D) - labbelledby wins - img with alt text and describedby - which wins? - doesn't seem to be covered in the current documentation - img with alt text, title, describedby, and labelledby - which wins?
- img with alt text (E) - img with alt text and labelledby (D) - labbelledby wins - img with alt text and describedby - which wins? - doesn't seem to be covered in the current documentation - img with alt text, title, describedby, and labelledby - which wins? Examples: Image with alt attribute <img alt="foo"> will result a text equivalent of "foo". 1) prepend a space. text equivalent is now " " 2) compute text equivalent A) Node is not hidden B) Node is not a text node C) aria-label is not present D) aria-labeledby is not present E) alt is a text equivalent provided by the native markup, so we add it to the text equivalent. The text equivalent now = " foo" F) <img> does not support user entered content G) Text equivalent is not empty H) Text equivalent is not empty 3) append a space. text equivalent is now " foo " 4) trim whitespace, condense interior whitespace. [??? why did we add it and then trim it??? What happens to crlf? I think IE preserves them]. text equivalent is now "foo" Image with alt="" <img alt="foo"> will result a text equivalent of "". 1) prepend a space. text equivalent is now " " 2) compute text equivalent A) Node is not hidden B) Node is not a text node C) aria-label is not present D) aria-labeledby is not present E) alt is a text equivalent provided by the native markup, so we add it to the text equivalent. However, it's null, so the text equivalent is still " " F) <img> does not support user entered content G) Text equivalent is not empty H) Text equivalent is not empty 3) append a space. text equivalent is now " " 4) trim whitespace, condense interior whitespace. [??? why did we add it and then trim it??? What happens to crlf? I think IE preserves them]. text equivalent is now "" (empty string) [??? by this algorithm, alt=" " and alt="" are the same. ??? Image with alt and describedby [??? describedby not mentioned in text equiv algo. Shouldn't it fill in the "descritpion" field, rather than the name? 3.6.1.3 doesn't make sense to me... ???] <img alt="foo" aria-describedby="1"><span id="1">bar</span> will result a name of "foo" and a description of "bar". [should we be specfiying the algorithm, or just the results?]
Trying to pick apart the text equivalent computation to make it more understandable. The algorithm in this section is referenced in the next two sections on computing the name and description. Need to talk through items D, F, and G. D seems to be redundant with the second paragraph in 3.6.1.2. The "Results" section of F doesn't make sense. It's referring to something that doesn't exist. I assume it means the examples above, not below. The rest seem pretty straightforward. The WAI-ARIA spec seems to explain this much better actually: http://www.w3.org/WAI/PF/aria/#nameref
From Cynthia: The explanation in the ARIA spec is more clear. We should either copy that one here, move it here, or remove it from our document. I'm not sure I agree with the idea of the user-entered value of a form element contributing to the name... That should be the value in a name/value pair, not part of the name. That's what IE does. Does FF add the user-entered value to the name now? Let's make a new bug on that, and close this one using the spec text. [Created bug 7634] Resolved to replace current UAIG text with text from WAI-ARIA spec. Michael, does this need to be in both places or can we remove from the WAI-ARIA spec?
Related discussion took place beginning at http://lists.w3.org/Archives/Member/member-pf-editors/2009OctDec/0018.html Question of whether to sync between spec and UAIG was added as a Tracker issue 377 http://www.w3.org/WAI/PF/Group/track/issues/377.
The text equivalent computation may be removed from the spec and replaced with a reference to the UAIG. In any case, the section is not that large and can easily be kept in sync manually. This will not be done for now.