W3C

– DRAFT –
ARIA APG meeting

04 May 2021

Attendees

Present
bryanG, carmacleod, garaventa, isaacdurazo, jamesn, Jemma, jongund, MarkMccarthy, Matt_King, sarah_higley, Siri
Regrets
CurtBellew, sarah
Chair
Jemma
Scribe
MarkMcCarthy

Meeting minutes

<Jemma> Meeting?

<Jemma> Meeting agenda: https://github.com/w3c/aria-practices/wiki/May-4%2C-2021-Agenda

APG Redesign Project update from Seth and Isaac from Bocoup

s3ththompson: the big update is that we met with WAI last week, the ops team, and a couple other reps from education and outreach

s3ththompson: we talked about a plan to come up with a long term migration of APG content to a more readable site

s3ththompson: WAI said they came up with a minimal template that should be really useful for us

s3ththompson: we also talked aout the structure of APG in a different format. in order to not disrupt much, we thought we'd publish the APG as an unlisted URL under WAI, and we can use that as a sandbox

s3ththompson: APG repo would be unchanged, publishing unchanged.

Matt_King: a note about the design template - it's a page tempalte that'll be used for all the other WAI supporting resources for WCAG

Matt_King: it's basically the header and footer that'll be used by understanding WCAG SC etc., and they're right at the point where they want input

Matt_King: we're, as APG, definitely open to making suggestions

s3ththompson: i was happy to hear that too

Isaac: happy to meet everyone! i've been working with the ARIA-AT TF before, glad to be on the redesign project too
… I'm going to be doing some design research to get some ideas from you all

<Jemma> some design template Seth mentioned https://w3c.github.io/wai-wcag-supporting-documents-redesign/2020-07-15-prototype-act.html

<Jemma> https://w3c.github.io/wai-wcag-supporting-documents-redesign/

Isaac: I'd like to start with a series of conversations with you all; get some ideas, learn some assumptions

Isaac: Goal is to improve the architecture of the APG and help it become more legible and user friendly - especially for web devs

<Jemma> ACT template example - https://deploy-preview-51--wcag-act-rules.netlify.app/standards-guidelines/act/rules/html-page-non-empty-title-2779a5/

Isaac: Over the next couple of weeks, i'd love to have a session with anyone/everyone who is available, about 45 min, and we can talk about what you see as success and goals

Isaac: less formal interview, more casual conversation, so don't be shy!

Isaac: I'll send an email with more details right after today's call

<Jemma> + 1 to follow up email by Issac

Isaac: Questions?

Jemma: what's the milestones for the project? Do you know a rough timeline?

s3ththompson: we'd like to start writing some scripts to generate WAI's templates within the next month. hopefully put it on an unlisted URL so we can look at it

s3ththompson: Matt_King, you were saying a longer term migration would happen after the next APG publication?

Matt_King: we want to take the most iterative approach, so the goal is that as soon as APG 1.2 is done, that'll be the last time we publish to TR directory

Matt_King: from then on, we'll maintain the Github pages through the start of the project, will still look like what TR looks like

Matt_King: i expect we'll take all the time we have between then and ARIA 1.3's recommendation to iterate the design of APG and what'll eventually appear on WAI's site

Matt_King: once we fully support ARIA 1.3 then we can debut a fresh new look

Matt_King: it could be we're ready for the new design prior to that and that could be our new editors draft, but we can think about that later on

Matt_King: always hard to know how these things shake out

jongund: about scope - is the stuff on the WAI website just our stuff? or are WCAG techniques etc. going to be there too? in other words, will it still remain separate?

Matt_King: for now, phase 1 is getting skeleton together, phase 2 will be the first redesign. we want to think about fitting those things in our architecture, but that's more involved

Matt_King: the only integration we plan by the end of the year is the ARIA-AT support tables

jamesn: maybe the next part is links to other things in WAI to make it easier to find

Matt_King: yep. THEN we'll think about integrating tutorials etc.

Matt_King: the education group is keen on that, but that might be a 2022 thing

Matt_King: ANYONE who has any interest, please carve out some time and meet up with Isaac!

Isaac: I'm hoping to start tomorrow, and go for a couple weeks. I'm flexible, reach out!

carmacleod: we talked about (deep)linking into APG from the ARIA spec last week. question is why haven't we done that before? we figured it was generally a good idea

carmacleod: i'd like to make the links more deterministic in APG; more direct into the naming guidance table

Matt_King: currently, if you have no-toc on something, that means that you can't put an IDREF for respec on it - it'd break

Matt_King: or anything below an H3

jamesn: i can talk to you about that, we can work on something that'd be workable, like aria-practices-definition for example

Matt_King: whatever we do, we'll want to think about how we maintain that. we've had that problem for a while with not enough hands. glad to have jamesn's help

jamesn: it's always been _possible_ but there's a lot of ways to go about it so it's kinda complicated

Matt_King: I've always wanted to reference the keyboard docs from a pattern page

jamesn: the patterns pages aren't respec docs, so respec can't generate links. that'd be harder to do

jamesn: tell you what, let me take a look at all this and we can see what we can do, i'll talk to Marcus

Matt_King: it'd be good to have a stable way of referencing any heading or table

jamesn: so any place you want to externally refer to something, it'd need to be exported

Matt_King: okay, let's not get too deep.

jamesn: we can talk to the maintainers of the repo to see what we can do

Matt_King: let's have an issue opened to discuss this, referencing APG resources from specs

Matt_King: can someone open that?

carmacleod: i'm on it

Matt_King: thanks!

Matt_King: can you assign that to s3ththompson? s3ththompson, you can hand it to whoever needs it. please assign me too

Slider example update

Slider example (comments on temp, media seek, multithumb slider and rating slider example)

Jemma: lots of comments, and some reviews needing done

<Jemma> https://github.com/w3c/aria-practices/pull/1864

<Jemma> https://github.com/w3c/aria-practices/pull/1864#issuecomment-832123780

Matt_King: i think this one is close to ready, but i ahve one question on value-text

<Jemma> value text - degree symbol

Matt_King: it's currently using the degree symbol --

jongund: no, it's not. the value text is using the word degrees and spelling Celsius. the visual label is the way you're describing

jongund: may be a bug then

jongund: nope, aria-valuetext is using "degrees Celsius"

Matt_King: maybe i was looking an out of date one

jongund: the visible label, which isn't part of the slider proper, uses the symbol and abbreviation

Matt_King: okay, a non-issue then?

jamesn: no you're right Matt_King, the static HTML page doesn't have the update

<Jemma> https://raw.githack.com/w3c/aria-practices/slider-temperature/examples/slider/slider-temperature.html

jongund: i'm not seeing that though

Matt_King: may be a technical/githack issue, no biggie

<Jemma> media seek slider - https://github.com/w3c/aria-practices/pull/1863

Matt_King: at least at Facebook, we tend not to use symbols in labels or valuetexts. not sure if that's a wider standard

<Jemma> https://github.com/w3c/aria-practices/pull/1863#issuecomment-831958339

Matt_King: in the seek slider, are we spelling out hours etc.?

jongund: yep

Jemma: siri - your comment?

<Jemma> https://github.com/w3c/aria-practices/pull/1863#issuecomment-831958339

<Jemma> "NVDA/Chrome in browse mode:

<Jemma> Narrates the time duration text as real time. For example: 1:00 is narrated as "one o clock" instead of 1 min, 0:00 as "midnight".

<Jemma> The slider is not narrated as interactive (i.e. clickable) element. It is narrated as Media Seek slider 3 Minutes 36 Seconds of 5 Minutes. When I remove role="none" from the it is narrating it as clickable but this approach adds other issues like narrating the whole graphic as clickable. I checked native HTML input type="range" and the screenreader announced it as slider. So I am not worried with the current approach. Please let me know if

<Jemma> you won't agree with me."

siri: the actual text, the valuetext, is being read correctly. but in browse mode it's reading as a time

Matt_King: i think that's a synthesizer thing

+1 Matt_King

jongund: i think what's happening is the visible label is #:##, so it's probably picking it up like that

Matt_King: like when you first tab into it, oh okay. does it say that when you change things Siri?

<Jemma> https://raw.githack.com/w3c/aria-practices/slider-seek/examples/slider/slider-seek.html

siri: no, when I change it, it seems fine, only when i tab into it

jongund: so that's not part of the slider proper, but the visible label. not sure we can do anything

Matt_King: i'm not worried about it

Matt_King: i think that's an NVDA thing

jongund: and all screen readers would see that as technically separate from the slider node

<Jemma> https://github.com/w3c/aria-practices/pull/1758

Matt_King: so on PR 1863, we still need reviews done. working on mine, sent an email to Jess about hers. we still need the a11y review done by Siri

Matt_King: are you ready to approve?

siri: yep

Matt_King: to be clear - this is all for the media slider

Jemma: the only reviews left are Matt_King and Jess

Matt_King: I think this is where we're at with the temp slider as well

jongund: yeah, so you can take a look as part of the overall review

Jemma: back to PR 1758

Matt_King: vis design review - carmacleod what do you think?

carmacleod: visual design? aha, okay

Matt_King: oh, wait you approved

Jemma: that just leaves editorial and code by Matt_King and Jess

Matt_King: good to go on that one then

Jemma: PR 1867

<Jemma> https://github.com/w3c/aria-practices/pull/1867

Jemma: I added WCAG touch target info

<Jemma> siri's comment - https://github.com/w3c/aria-practices/pull/1867#issuecomment-832134630

siri: i'm seeing that we're adding information. if I tab to it, I see "none of 5 stars" etc. but if I lose focus or browse, it doesn't say that anymore. it only says "none"

Matt_King: so how did you change the value?

siri: If I change the mode to browse, and change the value, then tab out or change focus, it seems like the aria-valuetext doesn't update

Matt_King: so I just changed it, then went into browse mode, and it's telling me the correct information

siri: but if you don't ever tab to it, the user never hears "out of 5 stars" or similar - whatever the range is

Matt_King: JAWS announces the min/max when in browse mode, NVDA doesn't

Matt_King: doesn't the valuetext string include the "of 5 stars" or similar?

jongund: when the page loads it says "none of 5 stars"

Matt_King: when the page loads, is the "of 5 stars" part of the value text?

jongund: when it gets focus then it adds the additional info. so when the page loads, it just gives the current state (like "no stars")

Matt_King: so I'm trying to figure out what triggers the "of 5 stars" to go away

Matt_King: it's reading the way we expect...

siri: so if you change the value, then go out and come back to the slider, it doesn't announce it.

jongund: when the page loads, the value of -valuetext is just "no stars"

Matt_King: so when i changed the value from half to one star, and changing NVDA to browse instead of focus, NVDA just says the value is one star

Jemma: seems like we might need more time. my testing with Android was hopeful, seems like it's getting better

Matt_King: so I think Siri's suggestion of when the value changes, or onBlur would be good. might as well do it onLoad too

jongund: i'll do that

jongund: do we have an update on JumpTo?

Matt_King: let's get that in a minute, let's go to the next agenda items

Add deprecation warning to Collapsible Listbox example

<Jemma> https://raw.githack.com/w3c/aria-practices/remove-listbox/examples/listbox/listbox-collapsible.html

<Jemma> https://github.com/w3c/aria-practices/pull/1852

Matt_King: i put some comments into the PR, I want to know if the warning's appearance and words are sufficiently obvious

Jemma: yes, and i was thinking about adding it to the spec

Matt_King: we can have the warning on the page, but since it's deprecated, i think we should remove all links to it

Jemma: +1

+1

jongund: +1

<Jemma> +1 for removing the example

jongund: there should be something in the index saying it's deprecated

Matt_King: oh yes

Matt_King: could we add a data attr to exclude something for deprecation?

jongund: maybe a metatag in the head

Matt_King: or on the H1, or... does it have to be anywhere specific?

jongund: can use a query selector, so it could be anywhere

jongund: this is in a branch right now, right?

Matt_King: yeah we'd merge first then do this

jamesn: as to removing links - instead of just removing the links, should we have a deprecated section of links?

Matt_King: opposition to adding "(deprecated)" to the page title?

jamesn: that's a good idea

jongund: could be useful for the index too

+1

Matt_King: so, jamesn, instead of removing from the index, we could have a section of deprecated content

jamesn: i'm picturing a dev who has followed this to a T, but mgmt comes back and says WTH - might be useful in that case

Matt_King: maybe we add the date too then?

jamesn: that'll be really important as we move forward

+1 jamesn's thoughts

Jemma: so lets add some comments to reflect jamesn's ideas

siri: how will the browsers or APIs handle it? will they be deemed inaccessible if they use deprecated things?

Matt_King: that's more of an ARIA question, not APG. this is just deprecating an example. in your example though, browsers didn't remove support of anything like that

Jemma: let's close the meeting, thanks everyone!

Matt_King: thanks friends!

Minutes manually created (not a transcript), formatted by scribe.perl version 131 (Sat Apr 24 15:23:43 2021 UTC).

Diagnostics

Succeeded: s/Seth: the/s3ththompson: the

Succeeded: s/wuold be unchanced/would be unchanged

Succeeded: s/th epage, but sincee/the page, but since

Maybe present: Isaac, s3ththompson