W3C

– DRAFT –
ARIA WG

17 February 2022

Attendees

Present
aaronlev, BryanGaraventa, cyns_, jamesn, Joanmarie_Diggs, MarkMcCarthy, Matt_King, RobertoPerez, scotto, valerie_young
Regrets
StefanSchnabel
Chair
JamesNurthen
Scribe
chlane

Meeting minutes

chlane present+

New Issue Triage

MarkMcCarthy: thanks

i know its a regex I just type slow

<MarkMcCarthy> https://github.com/w3c/aria/issues/1696

jamesn: agrees that a text child should label a descendant

jamesn: scotts recommends native behavior

they are coming up with a new select menu

authors need to add group labels

jamesn: should we not wait on outcome of those discussions

scott, open UI, do whatever, may ship without A11y

need more involvement or issues from scott

jamesn: need to talk with them, we should get a formal mechanism

jamesn: should not happen hap hazardly

jamesn: assinging to himself and 1.4

label as openUI

siri_: not understanding? why use role = group in listbox?

jamesn: not allowed on this usage, would need spec change

siri_: scott this is a reduced test case

jamesn: like the <optgroup>

scott we need to solve for groups of options

siri_: cannot use headings

roberto: any behavior examples?

jamesn: visible label needs aria-hidden

true

New PR Triage

all editorial not worry about them

Deep Dive planning

potential testing meeting

Bryan accname?

dissent between implementers?

bryan, further discussion was necessary

aaronlev: webkit didn't have aria description

aaronlev: macos 11 ax custom content

<aaronlev> https://bugs.webkit.org/show_bug.cgi?id=206253

before that it will be ax custom hint

jamesn: james c requested changes, none blocking

bryan will go through it with melanie

jamesn: review comments unresolved

jamesn: 8 comments from james c with no responses, some editorial

no meeting next week

Reminder - rejoin Group due to new charter

1.3 Blocking issues

1159 no updates

<MarkMcCarthy> https://github.com/w3c/aria/issues/1177

#1177 no update from Matt

Initial aria-textseparation (depends on generic PR being merged)

defer 996

Consider collapsing Base Concept category into Related Concepts

scott no update assigned it to himself

we are good on this one

Inconsistency between native and ARIA listboxes when implicit aria-selected is provided

needs reviews

<jamesn> https://github.com/w3c/aria/pull/1682

scott will review

Should ARIA provide better support for routing in single-page applications?

<jamesn> https://github.com/WICG/app-history#customizations-and-consequences-of-navigation-interception

scotts testing of WICG

app-history

with a screen reader

things were read out

no ARIA work

browser implementers should look at it

Matt, we look at a proposed change in a different spec

scotto: this could add announcements for routing

jamesn: it would allow a SPA to pretend it is reloading

transparent to the users

scotto: all the stuff we have to fake

Matt should ARIA provide a way for page authors to give guidance to screen reader to ovveride default experience?

commn problem

should we assist with ARIA

cyns_: volunteers to work on this

<jcraig> https://github.com/whatwg/html/issues/5562

jamesn: is this mandated? would an author have to change the way they use the app history api?

scotto: this is allowing authors to do more

scotto: reference to user report within the last year that incorporates what people are doing manually

<jamesn> https://github.com/WICG/app-history#customizations-and-consequences-of-navigation-interception

research by fable labs

scotto: to be clear, the whole point, the wcig proposal is linking this issue, a11y community should look into it

cyns_: volunteers, spoke with dominic

jamesc html proposal 5562 changing window.history

no new a11y API needed

scotto: it would be sufficient, ideally

could be an extention that ARIA picks up

could be brought back to ARI

A

jamesc assistive tech fallout could occur

may be that we don't want to do that if the web author changes focus then fires this or vice versa does that matter

bryan asking about tabs implementation using window.pushstate

to track active tab

focus doesn't change

jamesc the proposal wont break that use case

aaronlev: related to chrome issue

ff had back and forth still in memory just shows up no normal page load mechanism

the back/forward cache is turned off

matt related to windows tabs?

screen readerr is still reading previous tab until he hits escape

aaronlev: they did see that

Matt_King: in those cases, no focus change event & api nothing gets fired?

aaronlev: it is relevant to api

screen reader treats it like a psuedo page load

cyns_: add these questions to issue

jamesn: TPAC

A

10 plus in person

dates Sept sometime

9/12 - 9/16

jamesn: no announcements sent out

jamesn: chairs getting interest

Minutes manually created (not a transcript), formatted by scribe.perl version 185 (Thu Dec 2 18:51:55 2021 UTC).

Diagnostics

Succeeded: s/roberto any any/roberto: any

Succeeded: s/matt will review/scott will review

Succeeded: s/sudo page/psuedo page

Succeeded: s/rrsagent make minutes/

Succeeded: s/jamesn: TP/jamesn: TPAC

Maybe present: roberto, siri_