<jamesn> agendabot, find agenda
<agendabot> jamesn, OK. This may take a minute...
<agendabot> clear agenda
hi
am I on the wrong zoom call?
ah ok.
thank you, jamesn
zakim scribe pkra
<scribe> scribe: pkra
zakim next item
zakim: next item
james: core-aam, aria as
evergreen specs.
... non-chair opinion: core-aam seems good as evergreen. For
browsers, they should look at latest.
... but would like to have discussion
<jcraig> +1 for moving AAM to Evergreen... I see value in standard Rec track for ARIA, but could go either way.
james: for aria, we should maybe have a deep dive
mattking: evergreen is now a standard option?
james: yes, it's in the process docs now.
mattking: what's the process around it? e.g., 2 implementations.
james: still our process, i.e.,
for adding to main branch.
... deep dive seems to give better time box
mattking: for aam's seems uncontroversial. good starting point.
jcraig: I had proposed core-aam a
while back. Just makes sense, non-controversial.
... we have discussed it already several times. We now have the
process. Let's put it on the agenda to decide.
... then discuss ARIA.
jamesn: will schedule it for discussion.
jcraig: I meant: put it straight to email vote.
(pkra apologises)
jamesn: next week virtual
tpac.
... at our meeting time are two conflicting meetings
... feels like we shouldn't have a meeting at that time.
... we could move it an hour earlier (which also has conflicts
but less severe)
... no objections to skipping next week?
<MarkMccarthy> https://www.w3.org/2020/10/TPAC/registration.html
jamesn: none so we skip.
<jamesn> https://www.w3.org/2020/10/TPAC/breakout-schedule.html
<MarkMccarthy> oops - go back on epage from there
jamesn: nov scheduling. We might
skip Nov 19, and will skip Nov 26 for US Thanksgiving holiday.
(US thanksgiving)
... may skip Nov 19
mattking: this PR is now ready
for review.
... need input from authoring and browser side.
... from browser side, there's new normative language. We're
relatively sure it matches behavior.
... on authoring, biggest change is strong language to
discourage checked and selected at same time.
... but potential use cases are covered as well
... the summary is a good starting point.
james: we need reviewers!
jcraig: I'll review.
mattking: there's a lot in
#1052
... please take a look before reviewing the PR itself.
joanie: we should ask james teh, too.
jamesn: sarah's PR has enough
reviews. Ok to merge?
... => go
mattking: can merge practices change now
jamesn: we should do 1.3 triage
jamesn: start #907
... from Aaron.
... move to 1.4? Or just close?
jcraig: value structure does not
allow for time codes. even shorthand is ambiguous.
... we'd want to provide min/max in same time code value
... but seems 1.4
jamesn: move to 1.4
... next #908
... just to group the new elements (sup/subscripts)
melanie: abstract so not for authors?
jamesn: yes.
jcraig: could there be a complex descendant inside an element with the proposed "phrase" role?
mattking: the main thing from ontology is inheritance. If it would simplify some of that, then yes.
jamesn: would anyone do that
analysis?
... assigning myself.
stefan: if we need a text-related meta element in the ontology, then it sounds good. Cf phrasing content.
<StefanS> https://html.spec.whatwg.org/dev/dom.html#phrasing-content
jamesn: I'd favor 1.3 since it will be more complex later
stefan: span, link no concept for inner text.
jcraig: generic?
jamesn: just text node?
jcraig: our discussion today shows why people have are confused by the name "phrase".
mattking: it's abstract, spec not authors.
jamesn: right but if we're talking about child content, we end up putting it in front of users
melanie: are roles meant to give a way to parity with HTML or extra way to do the same thing?
jamesn: I see it as a mixture. Some are for role parity but some are not, they go beyond what HTML offers.
brian: historical perspective,
people created lots of silly elements, using spans and divs,
impacting accessibility. ARIA was invented to create a
mapping.
... recently, parity became a focus. And to some degree
confusion.
... it can be similar to what you encounter with native
widgets. But there are others where you can only realize it
with ARIA (e.g., some sliders)
<jcraig> abstract role prefix issue filed as https://github.com/w3c/aria/issues/1345
melanie: would it be helpful to differentiate retro-fit ones vs those that go beyond?
mattking: perhaps better fit for
practices.
... aria is actually host language independent.
... aria is not dependent on HTML so that other languages can
use it as well.
... we have a "related concepts" field.
melanie: for me "HTML = rendered in browsers"
<jamesn> https://w3c.github.io/aria-practices/#structural_roles
Sarah: html-aam might be the right area to look at?
jamesn: and structural roles
<sarah_higley> here are html-aam element/role mappings: https://w3c.github.io/html-aam/#element-mapping-table
<msumner> thank you!
jamesn: some of the role mappings
are "mostly" the same but not quite.
... number input type.
... spin button is much more than that.
... issue #910
mattking: does this question required owned elements?
pkra: I honestly don't recall. we can close this from my end.
<jcraig> https://github.com/w3c/aria/issues/910
stefan: a container is more robust than parent-child.
mattking: so a relationship
property rather than grouping.
... aria-owns is for that of course. but somewhat
backwards.
jamesn: I had the same thing when I first did it.
[pkra had trouble listening vs minuting]
jcraig: the reason to have group
was to map to e.g. nested list.
... so a nested container for tree items
... similarly for related roles.
brian: my understanding for the
original comment: in examples for trees, you often have a list
element
... in the LIs then you'd have children with the same
construct. Creates issues with scripting, instead LI would not
have role on it but then, say, a link inside as triggering
element, next to it a container with the same structure. Then
you'd have to work with aria-owns to connect.
Sarah: for me there's still a problem in the hierarchy
mattking: I just modified some language that is related.
jamesn: keep this open?
jcraig: pkra can you go back and find out what you had in mind?
Sarah: I still think grouping tree items might be legitimate.
<jcraig> You'll want to dismiss Zakim too I think
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/We should skip Nov 19, 26/We might skip Nov 19, and will skip Nov 26 for US Thanksgiving holiday./ Succeeded: s/preface/prefix/ Succeeded: s/could there be a complex descendant?/could there be a complex descendant inside an element with the proposed "phrase" role?/ Succeeded: s/shows why people have problems with the terminology./our discussion today shows why people have are confused by the name "phrase"./ Default Present: Irfan, Jemma, StefanSchnabel, Joanmarie_Diggs, harris, MarkMccarthy, msumner, carmacleod_, sarah_higley, pkra, Matt_King Present: Irfan Jemma StefanSchnabel Joanmarie_Diggs harris MarkMccarthy msumner carmacleod_ sarah_higley pkra Matt_King CurtBellew jcraig Found Scribe: pkra Inferring ScribeNick: pkra Agenda: https://lists.w3.org/Archives/Public/public-aria/2020Oct/0037.html 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: 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]