Re: ISSUE-540 -

Hi Joseph,

Apologies for the delayed reply.

The email thread starts at:

The issue was raised because of a change in how Webkit was mapping
@aria-labelledby and @aria-describedby, which was possibly a little
confusing were one to expect that AXTitle was for mapping the
accessible name, and AXDescription was for accessible description,
which isn't really how it works, hence the desire to clarify.

The relevant quotes from James are:

"One of the problems with abstracting different APIs into a single
lowest-common-denominator standard is that there will always places
where there is not a 1:1 mapping. For example, AXHelp is one of the
places where it's really more like tooltip text, so in the use of
describedby on an ARIA tooltip, it's a 100% match, but in other cases,
it's not as appropriate. "

"There are other differences between APIs besides the label
differences. Maybe some of the trickier API deltas could be listed
(including this example), but I don't think 'AXTitle vs AXDescription'
needs it's own section. "

In the end, we didn't get around to addressing such differences in the
html-api-mapping document beyond the way that Firefox uses MSAA

Yes, I would recommend that a section on differences such as these
between APIs be added to the core UAIG document as these are general
differences that would be inherited by the other mapping documents,
and so would be best placed in the core doc.

Hope this helps,


On Tue, Apr 8, 2014 at 6:51 AM, Joseph Scheuhammer <> wrote:
> Hello Jason,
> The AAPI task force recently came across ISSUE-540:  "Should the UAIG have a
> section to describe the API differences":
> This issue was assigned to the group on 24 Jan 2014.  We weren't sure what
> the issue is about, but did traced it to you via a link to an email thread:
> That thread dates back to 2012, and from what I can gather, it looks like
> the work was moved to the html-api-mappings document:
> But, I'm not certain.  Can you provide any insight?  In particular, is there
> anything that needs to be added to the core UAIG in this regard?
> Thanks.
> --
> ;;;;joseph.
> 'A: After all, it isn't rocket science.'
> 'K: Right. It's merely computer science.'
>              - J. D. Klaun -

Received on Monday, 14 April 2014 23:25:52 UTC