Meeting minutes
New PR Triage
<github-bot> I can't comment on that because it doesn't look like a github issue to me.
jamesn: Corrects conflicting language on ofcus managment needs ACT review
pkra: This was originally by Melanie, I redone this, Scott JamesN and JamesC reviewed it
… Not sure if we do need ACT review
… Because they didn't review
… Someone pointed out that the section has MUST statements but the individual rules have SHOULDs, so now the rules have also MUST
jamesn: I need to re-review it, it's so long time ago
jamesn: We should think if we need to do ACT
spectranaut_: Should we close the old PR?
pkra: I've set it up to do so
giacomo-petri: I can review for ACT, don't think it affects because there are no rules for this yet
jamesn: Please check the check box at the top and put not applicable. Thanks much
jamesn: Update defiintion of dropdown versus listbox rendering
spectranaut_: Add me so there's two reviewers
keithamus: I can do that oo
pkra: generate roleinfo after respec runs, please review
… Tried to reproduce as much breakage
jamesn: I'll try to get at this
WPT Open PRs
<github-bot> I can't comment on that because it doesn't look like a github issue to me.
spectranaut_: Anyone can review mine, it's small
giacomo-petri: I can review it.
spectranaut_: I'll email Simon
jamesn: Are the rest good?
spectranaut_: Yes
Deep Dive planning
<github-bot> I can't comment on that because it doesn't look like a github issue to me.
jamesn: I saw you're going to be at the hack fest
… The browser dev tools seems like a good topic
spectranaut_: Great idea
jamesn: Is it possible to schedule something at the end of the day?
spectranaut_: I can absolutely try to do that
keithamus: I'll be there too
aardrian: Surgery happening, can't attend personally
spectranaut_: It's in the middle of June, three days. It is breakout sessions and presentations, people trying to work together on things
keithamus: People refer to this as TPAC2, with same crowds and hallway tracks
jamesn: Not the easiest place to get to a Coruna
Daniel: Aria-actions is blocked.
jamesn: Is having a meeting going to help though?
spectranaut_: I cced Sarah lastweek, she said she'd get back to this soon
Matt_King: The challenges we were experiencing in the APG examples with the AT needing to be able to issue the click without the click generating side-effects I am not sure if we had consensus
spectranaut_: The takeaway was that Sarah was going to propose something morespecific for us to discuss
Matt_King: Hopefully we can get past that to make aria-actions feasible for sccreen readers
<Zakim> jcraig, you wanted to say aria-actions already shipped (behind a runtime flag) in Safari Tech Preview.
jcraig: aria-actions is already on Safari behind a runtime flag
… There's a path forwaard to keep it there, just we need to settle down on the interaction details
Matt_King: Does it work with the APG examples?
jcraig: It works but the APG examples are weird, I'd rather see some different examples
Matt_King: We have the other one we're working on. Actions on the elements of a list box
jcraig: That'd be a more expected example
<Zakim> benbeaudry, you wanted to say Edge will soon start implementation in Chromium
benbeaudry: Edge team is going to implement it soon
<aardrian> thumbs up
Daniel: Does it work with VoiceOver?
jcraig: Screen reader doesn'trequire any changes, this has been a native feature for years
clarify css generated content alternative text
pkra: Just a clarification. It made me look up what we have on our end and I thought this was a good one to talk with the CSS people but also as an aside it occurred to me that in reading accname is not easy to discern that the alt functionality of pseudocontent is taken into account
… Initially it was good for css-aam folks to take this on
Daniel: They were supposed to meet in March
pkra: They did review their pr and ask you to merge
melsumner: I think we should figure this out and put it in accname because there's other things were the spec is contradicting itself when it's talking about how the accname is built and generated in different contexts.
… Probably under step 2.6.2
jamesn: If we start to model every single place where CSS can change the ocntent it can easily explode.
… If we can push the content to CSS that'd be great
melsumner: Good idea to point to it
… We plan to do it for html-aam
keithamus: css-aam should be the place, right? CSS specs are a bit overwhelming?
jamesn: If the CSS specs have it that's fine, otherwise css-aam
keithamus: OK for accname to call out to css-aam
jamesn: I think so
bryan: I'd agree
pkra: css-aam -- I initially opened this is on which bases are we doing this? HTML and SVG acknowledge that there are host languages to ARIA, but that doesn't quite fit with CSS, but also there's not equivalent, and there's also CSS content 3 which talks about speech and that's not really relevant.
… What's the basis of this work?
jamesn: We talked about it with the CSS working Group a number of times
spectranaut_: Sure. But I don't know how to make it formal
aardrian: CSS is now starting to become a language in itself, should be good for us to formalize this in any way
<melsumner> agree with aardrian on this
Dainel: I'll open an issue, this should probalby be at a Charter level
pkra: I'll file a separte issue and assign it to melsumner
How should alt text be exposed?
jamesn: We'll keep it on the agneda for next week, as a task for people to get more familiar with this, unless people are prepared to talk about it now
melsumner: Is it really alt text for the Math element?
pkra: Yes
jamesn: All one word
Expectations for aria-hidden and focused elements
jamesn: There are a bunch of comments and commits on this, mainly on webkit. We did talk about it, but I kept it on the agenda because we had some comments by Apple
spectranaut_: I think the spec still needs to be updated
<Jacques> We do a version of this in chromium
jamesn: I think we have a couple of implementations, we just need to finish off this PR
spectranaut_: Some of the Chromium implementation was rolled back, right?
Jacques: I think we're good with this now
jamesn: Is this testable in WPT
spectranaut_: Yes.
jcraig: Rahim was working on some tests
… I'll file a new issue if the are no test
jamesn: Thanks.
… Rahim commented on the pass for clarity, maybe someone else can pick this up
jamesn: There is some more work to do on this based on JAmes Teh's comments
jcraig: There are two WPT issues but they're too broad, so I'll file a new issue pointing to this one and also linking to these two
CSS Carousels
Daniel: Those who were not here last week, please revisit the discussions around the explainer and my demos
Matt_King: I can take a look
aardrian: Same here
<spectranaut_> w3c/
[jamesn shows screen with demos]
Daniel: Please read the explainer and play with the demos
… In there, I compare CSS changes with ARIA patterns
jamesn: Shall we put this on the agenda?