accessible description computation revision #433
Conversation
This PR removes the various individual element accessible description computations and instead combines them into a single stepped computation listing, as apart from `summary`, `table` and `input type=button|reset|submit`, they were all the same. The revised accDescription still calls out these special cases (so nothing in the computation has changed here). We should probably discuss these again in the future. This PR also includes `aria-description` as being part of the first step in the accDescription computation (similar to how aria-label/aria-labelledby are mentioned as the first step in the element naming computations). AccName defines the priority of these two attributes in the accDesc computation. This PR is related to w3c/accname#69
index.html
Outdated
| <p>An <a data-cite="accname-1.2/#dfn-accessible-description">accessible description</a> MAY be provided to any HTML element that is a valid child of the `body` element.</p> | ||
| <ol> | ||
| <li> | ||
| If the element has an <a data-cite="wai-aria-1.2/#aria-describedby">`aria-describedby`</a> or <a data-cite="wai-aria-1.2/#aria-description">`aria-description`</a> attribute the <a data-cite="accname-1.2/#dfn-accessible-description">accessible description</a> is to be calculated using the algorithm defined in <a href="" class="accname">Accessible Name and Description: Computation and API Mappings</a>. |
There was a problem hiding this comment.
aria-description is not in ARIA 1.2 - so this needs to change to https://w3c.github.io/aria/#aria-description
There was a problem hiding this comment.
though, it seems that HTML AAM should probably just always link to the latest drafts of docs, rather than different versions of the specs
There was a problem hiding this comment.
I feel like this makes me feel like all of our specs should be evergreen? If you link from ARIA 1.3 to core-aam and then core-aam links back to the working draft? Seems odd to come back to a draft you didn't didn't start out in?
Anyway, for this, though, I agree.
There was a problem hiding this comment.
This PR certainly simplifies things and makes it a lot clearer.
I think you made a valid point with
it seems that HTML AAM should probably just always link to the latest drafts of docs, rather than different versions of the specs
Should the links be updated before approve/merge?
spectranaut
requested review from
spectranaut
and removed request for
jnurthen
There was a problem hiding this comment.
what do you think of these changes @scottaohara? when I first read you suggested change is stuck me as not as clear as the original table in the corrosponding accname pr.
|
@spectranaut appreciate the suggestions. i'm going to adjust just a couple of things, but yeh, that helps call out the element and its special condition better than as one lumpy sentence. |
attempt to simplify language taking a cue from @spectranaut's suggestions.
Co-authored-by: Valerie Young <spectranaut@gmail.com>
change to the editor's draft since it's not in 1.2 yet. will file another issue to change all links to editor's draft
change from the incorrect data-cite ref to use href=https://w3c.github.io/aria/#aria-description can change this to whatever the correct data-cite reference is when taking care of #437
|
merging this. This issue #437 will be where the spec is updated to link to the latest editors' draft of the ARIA specs |
This PR removes the various individual element accessible description computations and instead combines them into a single stepped computation listing, as apart from
summary,tableandinput type=button|reset|submit, they were all the same.The revised accDescription still calls out these special cases (so nothing in the computation has changed here). We should probably discuss these again in the future (for instance, see #278)
This PR also includes
aria-descriptionas being part of the first step in the accDescription computation (similar to how aria-label/aria-labelledby are mentioned as the first step in the element naming computations). AccName defines the priority of these two attributes in the accDesc computation.This PR is related to w3c/accname#69
Preview | Diff