Editorial: Should doc-biblioentry role be removed from dpub specs?

[giacomo-petri]

In the DPUB specifications, the doc-biblioentry role is marked as deprecated.

It requires an accessible name from author (aria-label or aria-labelledby) but does not support/inherit any ARIA states or properties (no aria-label or aria-labelledby?).

Additionally, a note clarifies that, according to ARIA, it should not be used because it is expected to be an accessible child of a list, which is not allowed.

Would it be reasonable to remove it from the list of roles entirely?

[daniel-montalvo]

Hi @giacomo-petri

thanks for putting this together.

The group decided to move these two specs forward to Proposed Recommendation. If we make these changes just now, this will mean it'll take longer for us to get to REC, which is what the DPUB team have been pursuing.

CCing @mattgarrish for his input on this.

[mattgarrish]

Yes, I think priority now should be getting 1.1 out now that we're at the PR stage. One reason 1.1 has taken so long to get this far is because we keep trying to address each new issue that arises.

Keeping the role as deprecated instead of pulling it entirely shouldn't affect usage, as least with publishers, as we've been warning them off it for some time now.

But this is definitely a good issue to keep on the radar for a next revision.

[giacomo-petri]

I'm fine with keeping it as is for now, but how should checkers handle this?
We're requiring an acc name from authors, but the specs don't allow aria attributes to be set.

[mattgarrish]

Sorry, I missed that part of your original comment.

I think I see what's happened now. The characteristics table for doc-biblioentry is missing its superclass role, so my guess is that the program that propagates the attributes doesn't know what to do with it and removes the row.

The role should still inherit the standard set of attributes for a listitem.

[pkra]

@giacomo-petri is there are a reason why this issue was filed in the ARIA repository and not the DPUB-ARIA repository? Otherwise I'd transfer it over.

[giacomo-petri]

My mistake, please move it to the DPUB-ARIA repository.
Thanks

[daniel-montalvo]

Triaged in 27 Feb meeting -- group thinks this is an editorial issue that should be fixed before publishing.

[aleventhal]

Knowing something is a biblio entry is useful for hardcopy Braille formatting.
But, if the author used doc-bilbiography with listitem children that could provide the same information.

[mattgarrish]

group thinks this is an editorial issue that should be fixed before publishing

To be clear, the editorial change here I assume is to add back the missing listitem superclass, not remove the role entirely at this stage, correct? I'll open a pull request on that assumption.