<scribe> scribe: MarkMcCarthy
Welcome Isabel Holdsworth!
jamesn: this one new issue today, a 1.2 or 1.3 issue?
jcraig: i don't think it needs 1.2, haven't seen aaronlev's PR yet, but.. 1.3 sounds better to me
jamesn: works for me
https://github.com/w3c/core-aam/issues/80
aaronlev: that works for me too
jcraig: seems like the answer to 1 might be not to expose them different, 2 we can discuss later/with 1.3
jamesn: add an agenda label?
jcraig: up to you, if you want to discuss in the meeting. expect it to be an easy fix or no fix
jamesn: sooner better than later - thanks!
jamesn: 2 new PRs, 1315 and and 1314
<jamesn> https://github.com/w3c/aria/pull/1314
jamesn: are we saying we don't want to use the old PR for this?
carmacleod: there was already one??
jamesn: it wasn't that simple, no worries. it was pushed and kicked down the road to 1.3
carmacleod: i'll still take care of it, adjust my PR accordingly after reviewing Wilco's PR, still needs 1.3
<jamesn> https://github.com/w3c/aria/pull/1162
jamesn: yeah, i don't want anything else in 1.2
<jamesn> https://github.com/w3c/aria/pull/1315
carmacleod: this may be
controverisal, but didn't hear objection to the SHOULD... so
let's get this done!
... anyone who is interested in landmarks at all, have a look
at this please
... the normative language was present in a few places but not
everywhere. the new-ish thing is that the language used to say
MAY, it now says SHOULD
jamesn: jcraig can you review?
jcraig: yeah, and i'm going to discuss within Apple as well
jamesn: can someone do the same for Google and Microsoft?
aaronlev: i can bring it to the right folks in Google - this is just for HTML?
carmacleod: yep
jamesn: okay, and Microsoft?
carmacleod: if not Sarah Higley, Melanie Richards?
aaronlev: well they have the same codebase as us...
jamesn: this is more the user interface though...?
jcraig: maybe Kevin Babbit?
aaronlev: he'd send it to the right person at leastd
jamesn: i can send an email i suppose
joanie: when i need info about UIA, he's who i flag for review, but i don't know him per se
jamesn: we could ask sarah higley to find out who the right person is?
carmacleod: that might be good
jamesn: i'll ping sarah and see what we can find out
aaronlev: maybe Alice Boxhall?
jamesn: could be, but we can't ping her on github. MichaelC?
MichaelC: maybe I didn't add her, so let me check
aaronlev: why is this an ARIA thing?
carmacleod: because of the change
in the spec about user agents. HTML issue is open and that's
being worked on. but it's a multipronged approach
... when i opened the HTML issue, they wanted a bit more in the
ARIA spec -- some of the definitions. this issue has been open
since 2016, so i thought I'd work on it
jcraig: aaronlev, the overlap is that landmarks were defined by ARIA and not HTML, while HTML works on a lot of mainstream changes. so it's kind of both
aaronlev: will this help users without AT?
carmacleod: essentially, the hope is such that it would
jamesn: this is only for implict roles? not explict ARIA roles?
carmacleod: right - UAs won't go there. i had to reduce the scope a little
jamesn: if we end up with that, seems like the ARIA equivalence might be worse?
carmacleod: or, use an HTML button! this would be the same type of thing "don't fake a main, use an HTML main"
jamesn: there's some exceptions, like an SVG experience
carmacleod: true.
jamesn: either way, seems like a reasonable approach that could be finessed later
Matt_King: seems like somebody splitting hairs - doesn't matter much on the code side.
aaronlev: my concern about having ARIA roles affecting UA behavior for those without screen readers MAY be a slippery slope
jamesn: it'd still be a big win to use native HTML
carmacleod: Matt_King, could you review, as well?
Matt_King: yep
carmacleod: i'd appreciate it, thank you!
jamesn: added you Matt_King, thanks
jamesn: deep dive for next week, second part of trees?
carmacleod: let's do it
jamesn: great! so next deep dive is next thursday one hour before this
jamesn: joanie and i discussed -
we want to try and put WR WD of 1.3 out near the beginning of
January
... so basically, we need to wrap before the holidays
... braille stuff is ready, annotations are ready, a few other
things too. we don't want to hold up WR if possible
... as we're planning, let's try to remember we only have a few
months to get 1.3 ready, and other things can push to 1.4
sina: +1
jamesn: if we miss something, another is coming
carmacleod: is there a theme?
Matt_King: theme is get it done [laughs]
jamesn: right - basically we just
want to get some new things finished up
... having something that people are invested in means we can
keep increasing the cadence
sina: how does that relate to 1.2's timeline?
jamesn: 1.2 is done, CR is done
(once some testing is done)
... 1.2 won't be "DONE" until we do WR WD of 1.3. doesn't mean
we shouldn't start getting reviews of 1.3
... danger is, people will look at 1.2, see 1.3 coming down the
pike, and maybe disregard 1.2
carmacleod: people don't tend to update tooling, recs, until it's at least in CR. or even wait for Recommendation.
jamesn: some tooling moves forward farily quick as we do (like axe)
sina: if i don't have to discuss comboboxes again, that'd be nice
carmacleod: yeah!
+1
jamesn: any other comments for this?
[silence]
jamesn: this is Wilco's issue
that carmacleod did a PR for before - Wilco asking for
updates
... answer - not a new issue in 1.2, this has been an ARIA
issue overall, so not treating as a blocker to 1.2
... working on it sooner than later, but not 1.2
... disagreements?
sina: well back to your original statement, we're trying to wrap 1.2, so...
jamesn: awesome
... scott had an issue about container roles and some things...
carmacleod is going to look at it and punt to 1.3
carmacleod: yep
jamesn: plan to get a PR into 1.3
sooner than later
... i'll leave the agenda tag on it, let's discuss next meeting
carmacleod
carmacleod: sounds good
jamesn: carmacleod is already
working on 350
... well to back up - i created projects to categorize some of
these. there are a lot of issues. but at least this should help
separate things out, and we can put some of these into those
projects
... 500. jcraig, does this need 1.3?
jcraig: i think it was a mistake in the first place, and needs fixing
jamesn: agree
jcraig: IIRC, joanie and Matt_King might know some of the reasons for objection, but i think this will be useful in general
jamesn: as long as theres an implicit semantic it'd be okay?
jcraig: yeah, but let's keep discussion for later, move it to 1.3
carmacleod: now that we have generic role, many more things have an implicit ARIA role, so...
jamesn: but roledescription is prohibited in certain circumstances
jcraig: i'm okay with an exclude list, but not necessarily an allow list
jamesn: 516 - this is editorial...
jcraig: assign me
jamesn: perfect
... 518 - asked for updates a year ago, i'm closing
jcraig: second
jamesn: 542 - aria-readonly for forms
Matt_King: getting this to WR by 1.3, who's championing?
jamesn: if Stefan was here, he opened it. i'll tag him on Github
jcraig: this would be a little difficult, unless all custom and no native elements
Matt_King: might be helpful to add that comment to the issue jcraig
jcraig: on it
jamesn: 559 - optional role
attributes for anchors... i've always thought something like
this could be useful
... i'd love to get this moving
Matt_King: why is it for the anchor?
jamesn: well a link OR button, in my opinion
Matt_King: that'll be the hard part of the discussion
carmacleod: final comment says it should be in HTML not ARIA
jamesn: sure, but it doens't mean we shouldn't/couldn't...
Matt_King: feels to me like this belongs in ARIA
jamesn: either way we'd want an equivalence. i'll take this one
Matt_King: hardest part is what to expect AT to do
jamesn: good point Matt_King, i'll put a deep dive tag on it
Matt_King: good idea
jamesn: 630 - deprecate
aria-flowto
... seems like it could be useful, but hasn' tbeen done well
and is fairly useless
jcraig: idealistically, i'm in favor of this, but I think Apple implemented portions of it in iOS
Matt_King: i think this is one
thing where we saw a problem and came up with an engineering
solution, but not with a sufficient understanding of real world
use cases
... might need more real world research
jamesn: i'd love to come up with a helpful replacement - there are real world problems that need solving. especially if iOS is successfully using it, or something like a flow chart
Matt_King: what aria-owns does is very inconsistent across UAs
carmacleod: aria-owns could work similarly, according to Alice
jamesn: this is about -flowto,
not -owns
... either way, jcraig are you okay with 1.3? you might be the
only one with an implementation
jcraig: works for me
Matt_King: yeah, user research will be really helpful
jamesn: 651, text formatting semantics. aaronlev, can you take this on for 1.3?
aaronlev left the meeting for today
jamesn: move to 1.4?
carmacleod: with a comment
jamesn: if no one wants to take it on, i'll move to 1.4
jcraig: he may care about that a lot, so maybe don't move it yet. it's important, but hard
Matt_King: yeah, this isn't a single issue, this is more like a whole project.
jamesn: 683, name from author/content
jcraig: maybe Isabel could followup with Steve?
Isabel: that's fine!
<carmacleod> https://github.com/w3c/aria/issues/683
jamesn: 684 - tri-state toggle; matt, you raised this
jcraig: i see this less with pressed than checked...
Matt_King: may indeed be editorial, was taken directly from the spec
jcraig: i've only seen mixed if you press other things. the "mix" is implicit based on other selections
Matt_King: just like a tri-state checkbox
jamesn: you're already assigned, do you wanna keep it? 1.3 okay?
Matt_King: yep, this should be pretty light weight.
jamesn: it's top of the hour! thanks everyone, remember to discuss trees next week before this ARIA call!
Isabel: can anyone join?
jamesn: absolutely! anyone from WG can join any meeting
Isabel: cool, thanks!
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) Succeeded: s/ask sarah/ask sarah higley/ Succeeded: s/maybe Kevin?/maybe Kevin Babbit?/ Succeeded: s/discuss with apple/discuss within Apple/ Succeeded: s/keep it 1.3/but let's keep discussion for later, move it to 1.3/ Succeeded: s/dicussion/discussion/ Succeeded: s/sounds editorial/matt, you raised this/ Succeeded: s/if not Sarah, Melanie?/if not Sarah Higley, Melanie Richards?/ Present: MarkMcCarthy pkra Joanmarie_Diggs carmacleod James_Craig Matt_King IsabelHoldsworth MichaelC Regrets: HarrisSchneiderman JonGunderson ScottOHara GretaKrafsig CurtBellew Found Scribe: MarkMcCarthy Inferring ScribeNick: MarkMcCarthy 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]