Meeting minutes
<Jemma> I am joinng now
jamesn: 5 new issues, skipping the first (#96) as its on agenda today
jamesn: 1382, 1.2 roadmap - not sure how to answer this on github but I'll take care of it
jamesn: 1381, followup on 1100. jcraig
<Jemma> https://
<MarkMccarthy> s/1100. jcraig / 1100. jcraig, is this a 1.3 or 1.4 issue?
jcraig: it could be 1.3, but ir eally needs input from joanie and aaron, maybe carmacleod. basically i took this up, just haven't filed it yet
jcraig: might be a nonissue, but not a 1.2 thing
jcraig: either way, this isn't pressing
jcraig: I assigned it to joanie
jamesn: 87 is editorial, 1.3
jamesn: 1380, we talked about it last week. adding 1.3 milestone, would be good to get it done
jamesn: only one new PR, just editorial. no need for reviews
jamesn: we have a bunch of things for the next deep dive - any proposals or anything?
jamesn: we talked about user actions for the web today, so i'll remove that one
jcraig: no immediate preference on which topic to go over
Matt_King: i thought on the accname one, we agreed that the intent of the spec was clear and bryan was going to propose modifications to make at least that example work
jamesn: there's something on the agenda today for that...
Matt_King: at any rate, the issue of self referencing labels, the algoritm, and spec example are all not exactly agreeing. it probably isn't a good candidate for a deep dive just yet
jamesn: i don't think there's anything to specifically discuss with that...
Matt_King: so maybe no deep dive next week, jamesn!
jamesn: sorted!
ACCNAME Suggested simplification<https://github.com/w3c/accname/issues/96 >
jamesn: we had this isue with accname, aaron proposed a simplification - linked above - to reorder accname slightly
aaronlev: 2c basically says if 2e applies, do that, otherwise go back to 2c and 2d
aaronlev: so, if 2e comes first, then that order makes a bit more sense. it seems like that what's chrome does anyway. so it just makes it easier to read
Matt_King: one question - i see references to the steps, people refer to the number and letter in issues/blog posts/etc.
Matt_King: because they're numbered and letter, rather than named based on what it does, when it's reordered I wonder if it'll cause confusion. not a reason not to do it, but it makes me wonder if it's truly that simple
bryan: it's kind of unavoidable though, like if other things need to be added, the order is going to change no matter what
jamesn: maybe we do away with numbering in the future, to avoid the issue altogether
jcraig: +1
Matt_King: that's what I was thinking
jcraig: might be better to put that in a separate issue though
Jemma: aaron talked about recursion concept for this issue too right?
aaronlev: i didn't rewrite everything, but i did call it descendent recursion, since that's what was intended
jcraig: so we have a short unique name for each step?
jamesn: makes sense
Matt_King: yes
aaronlev: i had been working on a better description of how to do description calc in a table format, people liked that. so maybe that's a good method too
bryan: that'd be awesome!
<Jemma> aaronlev: short unique name is recommended. it would be more appropriate to use the word "description calculation or decendants recursion"
jcraig: an example would be great!
<Jemma> rather than " recursion"
aaronlev: makes it easier to sort through so much information
jcraig: so if you have something, even incomplete, add it to an issue so bryan or whoever wants to can look at it
bryan: that would certainly be make things a lot easier
Jemma: makes me wonder if the table format would be 1:1 mapping, since it's not a flowchart? but i'm curious
aaronlev: i'll send an example, i promise it's easier than a flowchart!
carmacleod: back on the 2c/2d etc. PR, watch out for the IDs and fragment identifiers
jamesn: someone else can probably write this PR, so don't worry too much about it aaronlev
jamesn: i'll assign this to me. any objections any of this?
bryan: sounds great, i'm all for simplification.
Listbox and tree: clarify requirements for selected and checked<https://github.com/w3c/aria/pull/1340 >
jamesn: so this is all ready to go in but I didn't yet because I wanted aaron to look at it. i'd love James Teh and Matt_King to add some comments to aaron's comment
Matt_King: i thought I already did, but --
aaronlev: this is a response to your response, Matt_King. so basically it's not about the initial serialization, but we have to recompute the tree if user actions intervene
aaronlev: what i want - once it's made a decision to use the selection follows focus rule, it can stick that rule on that tree/list and not have to recompute the rule
aaronlev: it'd only make a rule if theres at least one item in there
Matt_King: if there's one item and the author inends to use checked but it's not declared, it won't work right until it's added
aaronlev: essentially yes
aaronlev: basically i just don't think it should be going back and forth and have to recompute etc.
Matt_King: makes sense.
jamesn: potentially we should get new reviewers/re-reviewers, might be helpful. jcraig, your thoughts?
jcraig: i can re-review if need be
Matt_King: doesn't have to be now, but if this is made explicit in spec, we would still need your review
jcraig: so if/when there's a new change, I'll definitely take another look
jamesn: aaronlev, okay to merge now and a new issue for future clarification?
aaronlev: definitely! no worries
jcraig: that'll make the second PR easier to review too
Matt_King: so we'll need an issue documenting when that new PR is there
Updated aria-setsize and aria-posinset to clarify usage for authors<https://github.com/w3c/aria/pull/1332 >
https://
we will wait for Melsumner's response.
1.3 triage<https://github.com/w3c/aria/issues?q=is%3Aopen+is%3Aissue+milestone%3A%22ARIA+1.3%22+no%3Aproject+sort%3Acreated-asc >
car: #350
… there was inconsistency and I fixed it. it would be great someone, JamesC, can review.
#966, we need to editorial work for this, "implicit value"
<jamesn> https://
975 is related to https://
https://
james: 953 can moved to aria 1.3
car and matt agree
#979
jamesn: no objection moving #979 to 1.4
#982 is editorial
#989
… add it to agenda item so that we can learn more about it.
<jamesn> https://
<jamesn> https://
this can be deep dive topic.
we can also invite other visualization experts, not only leonie
<carmacleod> https://
https://
(group discussed about this issue casually)