Accessibility Guidelines Working Group Teleconference

26 Sep 2017

See also: IRC log


AWK, KimD, kirkwood, jasonjgw, brooks, Roy, Melanie_Philipp, steverep, Laura, MichaelC, glenda, alastairc, chriscm, Mike, Elledge, Katie_late, Mike_Pluke, MikeGower, Katie_Haritos-Shea
EA_Draffan, Detlev, Jake, JF, Rachael, Kathy, Lisa, Denis, Andy_Heath, Bruce_Bailey, Makoto
alastairc, gowerm


<AWK> regets+ Jim_Allan

<AWK> +Joshue

<alastairc> hmm, did the password change?

<alastairc> ah, it's in the page, got it

<Mike_Elledge> Do we have a new meeting number?

<alastairc> scribe: alastairc

Registering for TPAC (https://www.w3.org/2002/09/wbs/35422/AGWG_TPAC2017/#wbsq4 – question 4 only)

<steverep> Do we need to register if we intend to come remotely?

AWK: An old item, 13 people signed up, 7 are likely to come, others less certain. Main purpose of this item is to remind people to register. We've had less people registered than we'd expect.

<steverep> Thanks :)

AWK: NB: don't have to register if you are attending remotely.

MichaelC: When you complete the registration there should be a link to pay. The payment due date is October 6th (ish), and the payment goes up after that day.
... also, hotel costs are going up a lot, they are selling out of the pre-booked spaces, could be $400/night or more. Book now!

SC Numbering

<MichaelC> https://lists.w3.org/Archives/Public/w3c-wai-gl/2017JulSep/1097.html

AWK: A few people offered to put together somethings based on last week's discussion. Seen a couple of emails.

MichaelC: Sorry I didn't send it sooner, but I sent an email to the list breaking it down. The proposal is not to change the numbering, or new numbering scheme. Less concern about keeping levels grouped together, so let them land how they land. Then, in quick reference we would provide ordering by level.
... we have IDs in 2.1, we should document how to use them, which is a combination of numbers and short-names.
... all that boils down to not making changes, and moving on.

<david-macdonald> https://docs.google.com/spreadsheets/d/1K-pHAu-7yfDiEuP0WbER1YldANYEBvqopxvJJH9mYbI/edit#gid=1482528443

<steverep> No one responded to mine - https://lists.w3.org/Archives/Public/w3c-wai-gl/2017JulSep/1072.html

AWK: Something the chairs and Michael have spoken about. There's some diversity of opinion, but the strongest thoughts seem to be around not changing those aspects.

<Glenda> +1 to WCAG 2.0 not changing at all. Deque AC position.

marcjohlic: What are the IDs in this context?

MichaelC: The part after the hash in the URL, which we will document more fully in 2.1.

<marcjohlic> Marc: for example "non-text-content" is the ID for 1.1.1

AWK: Some people have been using the numbers as IDs, the short-names have spaces. For 2.1 we've been generating those so it's very straightforward, e.g. Graphics contrast becomes graphics-contrast. As part of this, we'll take a more clearly linked approach.

David: We should probably explore a couple of things, e.g. I didn't see support for changing the triple-A criteria.
... the 4 tier numbering looked ugly, probably wouldn't work.
... I think Laura and others suggested reducing the visual reliance on the numbers, and then having the ordering follow the levels.

MichaelC: In terms of moving SC to keep conformance, there was a mockup sent around which people didn't seem to like.

<Zakim> steverep, you wanted to say this is not an actual solution and doesn't address merging, changing, etc.

MichaelC: numbers at the end are harder to do, I think it will make things harder all around.

steverep: Does this address any problems? If we want to move an SC from one to another, we can't do that. We can't change an SC from 2.0 to 2.1.

MichaelC: I think we're saying the changes we investigated turned out to be worse than the solutions. It does highlight some cosntraints. (missed a bit on moving things around)

AWK: Note there is one in the survey about moving an SC.

steverep: By doing it this way we've locked into not removing any SC, such as text-sizing which has public comments saying it is redundent.

AWK: That's something that would be revisited in Silver.

MichaelC: This came up from questions around merging SCs from 2.0 & 2.1, and highlighted a lot of the constraints.

Jason: I don't agree that numbering should be stable, or there was a promise of that. However, it isn't the largest issue we're contending with.

<laura> What happens if we need a WCAG 2.2?

Jason: but what happens when there is an aversion to change 2.0 at all. In a later 2.2 or 2.x, we have a doc we can't modify even with known short-comings. The constraints outlined mean we would need to move down the silver track instead of 2.x track.

<steverep> +1 to Jason - why we are deciding to lock ourselves in to every published SC as "perfect" is beyond my understanding

Jason: these and other issues means we will need to move to silver more quickly.

MichaelC: This proposal means building 2.2 in the same as 2.1, tacking on guidelines. It is getting cluttered looking, but it works.

steverep: If people are locked into the numbers and we continue with this, there isn't any point de-emphasising the IDs.

MichaelC: The point was not being disruptive. We'll encourage IDs use instead of numbers, but we can't stop them using them.

<steverep> Just assigning 1, 2, 3 as I suggested is shorter...

AWK: Some people like the numbers, and they are short, don't need translation.
... Can anyone not live with Michael's proposal. I.e. use existing numbers. Add new ones after existing ones (out of level-order).

Katie: Liked David's proposal, but can live with it.

<kirkwood> +1

RESOLUTION: Accept Michael's proposal on WCAG 2.1 numbering and IDs

Resolving Issues Survey (https://www.w3.org/2002/09/wbs/35422/resolvingissues1/results)

AWK: We have 6 issues, of those, numbers 1, 4, 5, 6 are all marked as accepted as proposed. As per WCAG 2.0 process, the survey is the best and earliest opportunity to comment. Therefore I'd like to call for the working group to accept those 4 as proposed based on unanimous consent.
... Please note we need to get through about 20/week, so discussion should be focused. These aren't big issues like accepting new SCs, these are responses to commenters. Any objection to accepting these 4?

RESOLUTION: Issues 154, 375, 157, 153 accepted as proposed

Issue 379

steverep: The accidental activation SC seems to make more sense under the 'operable' principle, we all seem to agree.

AWK: Note the comment in the survey was actually about a different issue, misplaced.
... This is a slightly more significant change, although not controversial.

David: So we just re-number it? I think it helps, it will be good for numbering overall.

AWK: If it moves, it will definitely have a different number. It does feel like it fits under operable better.

Michael: I don't think moving an SC is a normative change, so unless there's an objection?

Katie: I disagree it is just editorial, but agree with moving it.

MichaelC: Ok, need a CFC then.

AWK: Any objection to moving it? (CFC later)

RESOLUTION: Accept as proposed, Issue 379

Issue 373

MichaelC: Process concern: The issue was to move the word "easily", but don't think we should also change an "and" to "or", as people haven't reviewed the consequences of that change.

AWK: I had seen the change as editorial, 'postponing' and 'suppressing' are different things, time aspect different. But happy not to worry about that.

MichaelC: Would like to run it past Lisa who isn't on the call.

David: I think it's important for COGA to know something is 'easy' to do, but best tackled in the understanding rather than impact the testability.

<Mike_Pluke> +1

<Mike_Pluke> presenta+ Mike_Pluke

Jason: When talking about 'mechinsm being available', that isn't something the content author is always in control of, it could be from the user's side.

<Zakim> AWK, you wanted to ask if we are removing the definition

AWK: Do we need to remove the definition? I'm sure some people would object to that though. The concern is that changing from 'easily available' to 'available' , are we dictating the use of 'available' for the other 73 uses in WCAG? E.g. for info & relationships.

MichaelC: If we drop the term from the SC that content could be used in the understanding doc. Introducing that term does bring complications, so moving to understanding seems like the easiest path.

<kirkwood> agreed important to have Lisa weigh in

<KimD> +1 (remove "easily" as it's too subjective)

AWK: Any other comments? (None) Is seems we have general agreement on the call & survey to change the term "easily available" to "available", and it makes sense to move that content to the understanding doc.

steverep: I didn't realise there was a definition, and it is quite 'normative', maybe this needs to be explored to move the definition to the SC?

MichaelC: I thought about adding a note about being 'easily available', but it just didn't fit the normal pattern we generally use.

AWK: In looking at the definition, the 1st bullet is a GPII ref, has 'is reachable' as a fall-back, but actually required by the conformance rules by WCAG. It would have to be available on each screen.
... (James notes the steps in a process) ah, yes, that would also be possible.

MichaelC: I see why the term was included, but it doesn't fit with normative text, better in a note or understanding.

David: Any comments from the public?

MichaelC: Not yet.

JamesN: We've raised issues about the term 'easily' when we've found them, will do another pass through and would expect to raise it then.

AWK: We should have a concrete pull-request to see what it's like in a fully fleshed out form.

Mike_Pluke: The COGA TF thought that 'easily' was important, but it is clearly difficult in SC text, needs to be advisory.

David: Even with the definition? My previous proposal was to wait for public comments if we haven't got internal disagreement?

Mike_Pluke: I don't see how easily could be testable, great ambition, but not testable.

David: If we took out the word and put the definition in, would that be testable?

MichaelC: I'm not sure the definition would pass the testability bar, but I think we need a more concrete proposal in front of us.

Mike_Pluke: Originally I think the definition was part of it, but we moved it out.

David: Just as a mental exercise, I was curious to see if it would work.

AWK: We do need to be careful about the subjective terms, it is possible to do and make testable in some cases, but difficult to do.

David: Is there another term? E.g. prominent.

Jason: Just thinking about the implications. If we put that definition in normatively, then content authors would have to look at the available AT, and increase authors looking to create the mechanism if the definition has not been met. My personal opinion is that people creating AT are better at tackling this compared to authors.

MichaelC: We could have multiple proposals, we just need it in a particular timeframe.

<steverep> Concrete = cement (review of actual words) + aggregate (actual words/pull request to review)

AWK: For the group to say "this is the direction we should go", they will want to see what it looks like.

RESOLUTION: One or more concrete proposals needed for issue 373 - A mechanism is easily available

<scribe> scribe: gowerm

tracking progress on actions

AWK: Those 6 items completed because people took them on. Thanks! We need other people to take on issues. We have 122 issue open and handle 4 conclusively.

<AWK> https://github.com/w3c/wcag21/issues?q=is%3Aissue+is%3Aopen+label%3A%22AGWG+Work+item%22

AWK: THere are a few items unassigned, feel free to take. Send explicit email to WG when item is ready to review
... We send out the agenda 24 hours in advance, so please get them in in advance of that.

Steve: has to do with #372. The pull request is #393
... the accessible name one is also ready: #371

MichaelC: I will take on a few this week, including the Failures one

David: Question was 'should we have failures in 2.1?'

<alastairc> I have 9 issues assigned to me, I'll spend some time on them this week, they are mostly from WG members.

AWK: I would prefer to focus on older public comments
... The last link posted in IRC goes to a list of items we need responses to. For example, Issue #258 is available, from Animation from Interactions.
... This one is talking about vestibular issues.
... Wants to see at AA or AAA.
... What someone needs to do is look back and see what they are commenting on, and if the same concerns exists, and if they are likely to continue to exist.
... If resolved, then you point out what has been done to address concerns. If it still exists as a problem, it may be the signed up person needs to crystalize the issues and start a discussion on the mailing list

MikeE: How do we get back to the most recent version of the SC?

<AWK> https://www.w3.org/TR/WCAG21

AWK: This linked page at the top has a 'previous version' you can use to work back through each version.

Marc: #193 dealing with Orientation, it looks like something that can be addressed by Understanding, since it has been addressed in SC.

AWK: Write it up in response form so the WG can digest and make a decision.

<AWK> Understanding documents: https://www.w3.org/WAI/GL/wiki/Accepted_WCAG_2.1_SC

AWK: FOlks have committed to doing work on these things. We will start running through, asking for status.
... Our goal is we want to have a solid draft for each SC in place for TPAC
... If adding queue request, the structure is "q+ to say..."
... We agreed to a more casual approach to the incorporation of Understanding documents.
... You can make your own branch and create a pull request, or send an email asking for it to be merged

Steve: My question was about 'when' it will be pulled in.
... I heard AWK say he is going to wait for TFs. Right?

MichaelC: When people feel it's ready, that's when it will go in.

AWK: THey are all going to be merged in by TPAC unless they are completely not ready.
... Our intent is to have this initial round rethink the initial content in light of the current form of the SC.
... We will only get through the items if people step up and do the work they signed on for.

<laura> Bye.

trackbot, end meeting

Summary of Action Items

Summary of Resolutions

  1. Accept Michael's proposal on WCAG 2.1 numbering and IDs
  2. Issues 154, 375, 157, 153 accepted as proposed
  3. Accept as proposed, Issue 379
  4. One or more concrete proposals needed for issue 373 - A mechanism is easily available
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.152 (CVS log)
$Date: 2017/09/26 16:32:17 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.152  of Date: 2017/02/06 11:04:15  
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/link at top of irc has it//
Succeeded: s/MichaelC/David/
Succeeded: s/avialble/available/
Succeeded: s/with the current/in light of the current/
Default Present: AWK, KimD, kirkwood, Joshue, jasonjgw, brooks, Roy, Melanie_Philipp, steverep, Laura, MichaelC, glenda, alastairc, chriscm, Mike, Elledge, Katie_late, Mike_Pluke, MikeGower, Katie_Haritos-Shea
Present: AWK KimD kirkwood jasonjgw brooks Roy Melanie_Philipp steverep Laura MichaelC glenda alastairc chriscm Mike Elledge Katie_late Mike_Pluke MikeGower Katie_Haritos-Shea
Regrets: EA_Draffan Detlev Jake JF Rachael Kathy Lisa Denis Andy_Heath Bruce_Bailey Makoto
Found Scribe: alastairc
Inferring ScribeNick: alastairc
Found Scribe: gowerm
Inferring ScribeNick: gowerm
Scribes: alastairc, gowerm
ScribeNicks: alastairc, gowerm
Found Date: 26 Sep 2017
Guessing minutes URL: http://www.w3.org/2017/09/26-ag-minutes.html
People with action items: 

[End of scribe.perl diagnostic output]