<scribe> scribe: harris
<Greta> thanks
jamesn: sounds like a 1.3 issue to me
joanie: yea that is fine
jamesn: definitely 1.3 issue
https://github.com/w3c/aria/issues/1270
jamesn: is this (1270) 1.2 or 1.3?
joanie: if its easy cleanup then I'd say 1.2
jamesn: there is the topic of
multiselect comboboxes that sina has brought up
... some conversation has already taken place in the slack
channel which probably isn't the best place since not everyone
is there
... we should let sina voice his views on these
... exposing shortcut keys to AT is another topic
... another topic is: how NVDA is exposing dialogs and some of
the issues surrounding that. We probably want NVDA on that but
the 9am time is difficult for them so we probably want a
different time slot for them
harris: I'm super interested multiselect comboboxes
Jemma: is that topic specific to just NVDA? or could it be more broad to all AT?
carmacleod: if we're talking about checked/selected in comboboxes, perhaps we should discuss those as they related to tables and trees
<jamesn> https://github.com/w3c/aria/milestone/10
<jamesn> https://github.com/w3c/aria/pull/1261
aaronlev: I forwarded this to alice boxhall for review. This is more her area
jamesn: can we ask her when she might be able to get to it?
aaronlev: sure
https://github.com/w3c/aria/issues/1151
<jamesn> https://github.com/w3c/aria/pull/1224
aaronlev: aria-activedescendant
doesnt do anything unless it is on a focused element
... I found that our definition of "owned" is wrong
jongund: I'd prefer to not have that be part of this PR. It is a separate issue
jamesn: do you think this is already covered in another PR? A 1.3 issue
aaronlev: the definition of owned
makes it sounds like the ancestry chain can split
... there was another issue that should also be put into 1.3.
The definition of required owned elements - you have to say it
needs at least one but the example of a non-busy empty
listbox
... reading the comments - yea jon I think you've got the gist
of it
... the trick of using aria-controls to enable
activedescendant. In chrome, we don't enable that as loosely as
it says in the spec
jongund: right now chrome limits it to just textbox
aaronlev: yea but we don't go
that far up the tree to look for it
... the APG example with the empty listbox could be 1.3
... if the author uses aria-controls and aria-activedescendant
NOT on a textbox, then it will break
... we should either change the spec or change chrome
... I'd prefer to change the spec
jongund: we don't want to get
ahead of what browser do, do we?
... I don't think we really have a choice
aaronlev: we should meet half way
aaron: change the spec to say its only on textbox
aaronlev: and change chrome implementation
jamesn: so activedescendant in chrome requires the owned relationship?
aaronlev: the part that fires
focus events, doesn't care where its coming from
... the part that has the focusable state is different
<jamesn> https://github.com/w3c/aria/projects/13
https://github.com/w3c/aria/issues/712
aaronlev: I dislike
aria-relevant
... people speak up about it from time to time. Normally, live
region text doesn't get spoken when a node is removed. Say we
have a chat log where someone removes a message. Usually the
chat client adds a textual indication that it was removed
... the 2nd problem i have is: the difference between additions
and text is a bit unclear. Some browsers implement this
differently as far as how they handle dom changes
... nobody seems to use aria-relevant properly
jamesn: anybody object to adding this to 1.3
https://github.com/w3c/aria/pull/1072/files
jamesn: should I mark this as low
priority?
... possibly superseded by 712
... I've created a bucket dependent on other issues - I've
thrown this one there
https://github.com/w3c/aria/issues/322
<jongund> need to leave a little early today
jamesn: it'd be cool to trigger the browser to do what it does for a "real" page load event
greta: I'd prioritize this as
high because we use this a lot at washington post
... we refresh the news articles every ~30 seconds
aaronlev: if I'm on an app that is changing browser history. so if I use history back, it'll actually go through the history points
harris: isn't it possible to properly implement an accessible SPA today?
aaronlev: maybe browsers should hook into the browser history api and notify AT?
harris: +1 I like that idea!
jamesn: I put this on high priority. I put it as 1.4
greta: one of the good examples we have is: if we change which article has priority (visually), we don't have a good way to easily say "We've bumped this article up as the priority" or "This article is important right now"
https://github.com/w3c/aria/issues/72
aaronlev: its worth thinking about
https://github.com/w3c/aria/issues/671
aaronlev: theres actual use cases for this
jamesn: is there a spec change needed? or should we close?
aaronlev: seems like core aam thing, no?
joanie: this actually came up in
a number of things I've been working on. There are times when
the author is repeating the message and other times when it is
just even spam
... I don't think this belongs in core-aam
aaronlev: the author has no way of doing this without getting hacky
joanie: orca ignores duplicate
events
... this is meaty but let's add it to the deep dive list
https://github.com/w3c/aria/issues/1216
https://github.com/w3c/aria/issues/821
jamesn: can I put (821) that as low?
https://github.com/w3c/aria/issues/832
jamesn: can I deep dive this one?
harris: sounds like a good deep dive candidate
jamesn: I'd love to not have to use the old offscreen live region hack
https://github.com/w3c/aria/issues/1104
aaronlev: i still think log is useful
https://github.com/w3c/aria/issues/1017
+1 to high
rssagent, make minutes
rssagent, stop
This is scribe.perl Revision of Date Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00) Default Present: JamesNurthen, MichaelC, Joanmarie_Diggs, MarkMccarthy, harris, CurtBellew, Jemma, carmacleod Present: JamesNurthen MichaelC Joanmarie_Diggs MarkMccarthy harris CurtBellew Jemma carmacleod Regrets: BryanGaraventa PeterKrautzberger StefanSchnabel Found Scribe: harris Inferring ScribeNick: harris WARNING: No date found! Assuming today. (Hint: Specify the W3C IRC log URL, and the date will be determined from that.) Or specify the date like this: <dbooth> Date: 12 Sep 2002 People with action items: WARNING: Input appears to use implicit continuation lines. You may need the "-implicitContinuations" option. WARNING: IRC log location not specified! (You can ignore this warning if you do not want the generated minutes to contain a link to the original IRC log.)[End of scribe.perl diagnostic output]