Re: Another AccName sanity check: presentational role and name computation

Hey Jon.

I agree 100% with what you say here:

On 03/19/2018 03:58 PM, Gunderson, Jon R wrote:

> We need to make the algorithm simpler to help people understand it and not make mistakes.

But I think that ship has sailed -- at least insofar as 1.1 is
concerned. (As a reminder, we're blocking the specs of two Working
Groups.  AccName's got to ship.) I would really, really, really like to
simplify it for 1.2. That's one of the proposed topics for the upcoming
face-to-face. :)

Back to the problem at hand: If we all agree with my assessment -- which
is the assessment of both Firefox and WebKit -- we update the
expectations and have eliminated another case where we lack two passes.
On the other hand, if we agree with what Bryan thinks is the correct
outcome, then I think we need to fix the language so that it is
abundantly clear. But I cannot submit patches to user agents to "fix"
something which the spec (IMHO) says is not broken. So what I need from
y'all is a verdict so that I can move forward one way or the other.

Thanks in advance!
--joanie

Received on Monday, 19 March 2018 20:10:31 UTC