Re: accname-aam need to clarify aria-owns in Step 2F (github issue #538)

Hi, Joanie. Catching up with old emails, sorry for being slow on this.

My understanding is aria-owns affects on the accessible tree, thus it'd be
reasonable to expect aria-owned children in name computation results I
think.
Cc'ing Aaron for his opinion, since he was around the name computation
stuff lately.

Thank you.
Alex.


On Thu, Mar 23, 2017 at 1:51 PM, Joanmarie Diggs <jdiggs@igalia.com> wrote:

> Hey Dominic (and Alex):
>
> When you get a chance, could you please look at this issue:
> https://github.com/w3c/aria/issues/538
>
> My understanding from today's meeting, and from Joseph Scheuhammer, is
> that the intended meaning of "child node" in Step 2F is "DOM child
> node." The spec just fails to make that explicit. Firefox is using that
> interpretation, but Chrome is not. And since you were the one to make
> the change in Chrome [1], we wanted to get your insight and opinion.
>
> If you feel that the DOM-child-only interpretation is reasonable, we'd
> like to make that change to the AccName AAM 1.1 spec prior to our
> transitioning it to CR. On the other hand, if you feel like your
> implementation is the right way to go, we need to discuss it and reach
> consensus.
>
> Thanks in advance for your time.
> --joanie
>
> [1] https://bugs.chromium.org/p/chromium/issues/detail?id=561766
>


On Thu, Mar 23, 2017 at 1:51 PM, Joanmarie Diggs <jdiggs@igalia.com> wrote:

> Hey Dominic (and Alex):
>
> When you get a chance, could you please look at this issue:
> https://github.com/w3c/aria/issues/538
>
> My understanding from today's meeting, and from Joseph Scheuhammer, is
> that the intended meaning of "child node" in Step 2F is "DOM child
> node." The spec just fails to make that explicit. Firefox is using that
> interpretation, but Chrome is not. And since you were the one to make
> the change in Chrome [1], we wanted to get your insight and opinion.
>
> If you feel that the DOM-child-only interpretation is reasonable, we'd
> like to make that change to the AccName AAM 1.1 spec prior to our
> transitioning it to CR. On the other hand, if you feel like your
> implementation is the right way to go, we need to discuss it and reach
> consensus.
>
> Thanks in advance for your time.
> --joanie
>
> [1] https://bugs.chromium.org/p/chromium/issues/detail?id=561766
>