W3C

– DRAFT –
AGWG Teleconference

01 March 2022

Attendees

Present
alastairc, Azlan, bruce_bailey, Chuck, Francis_Storr, garrison, GreggVan, JaeunJemmaKu, janina, jaunita_george, Jen_G, Jennie, JustineP, Katie_Haritos-Shea, KimD, kirkwood, Laura_Carlson, Lauriat, MelanieP, michael, myasonik, Nicaise, sarahhorton, shadi, ShawnT, StefanS, SuzanneTaylor, ToddL, Wilco
Regrets
JakeA
Chair
alastairc
Scribe
Jem, shadi

Meeting minutes

Timezone warning

<AWK> +AWK

<AWK> March 14 is US time change

AC: coming up to time changes
… US changing first
… meeting on US time, so no change for US folks
… but for everyone else
… no meeting on 15 March

Introduction for WAI groups to a proposed Legal Entity for W3C https://lists.w3.org/Archives/Public/w3c-wai-gl/2022JanMar/0116.html

AC: W3C aiming to transition to a single Legal Entity
… session coming up with the Advisory Board member Leonie
… during an upcoming APA WG meeting
… everyone in this group also welcome

WCAG 3 Protocols and Scoping Next Steps https://docs.google.com/presentation/d/1b5xHQWBzoYdKp7BfPgIUBCpz-yaDOx_kSq_HlQxcFh0/

<GreggVan> /www.invaluable.com/auction-lot/original-1920s-cappiello-king-liquor-poster-veuve-44-c-b7630c45a5//www.invaluable.com/auction-lot/original-1920s-cappiello-king-liquor-poster-veuve-44-c-b7630c45a5

RMB: [slide 2] a way to frame conversation

[Rachael talks through slide 2]

<Jem> +1 to the way managing presentation session.

RMB: [slide 3] learning from both WCAG 2 and WCAG 3, no perfect start

RMB: [slide 4] different sub-groups working on different aspects

RMB: [slide 5] work to be seen in the context of the active work of the sub-groups
… how these different aspects fit together

RMB: [slide 6] looking at what it is that we are testing

RMB: [slide 7] think there are 4 types of tests
… first is constant (objective) test

RMB: [slide 8] second is a condition (subjective) test

RMB: [slide 9] third is a new type of test, not yet in WCAG
… not set by WCAG, it is set internally

RMB: [slide 10] fourth level of testing is protocols
… even more abstract than test case
… more ambiguous and difficult to define

RMB: [slide 11] mapping the 4 types of things we tested and the 4 types of tests
… matrix where automation moves expectations to the left over time

RMB: [slide 12] explored with alt-text as a specific example

RMB: [slide 13] possible next steps if we decide to work in this direction
… trying to create framework for a shared approach

<Ryladog> Brilliant!!

<michael> Nice summary

<Lauriat> +1, thank you for running through that, Rachael!

GV: not sure I fully understood all 4 types of tests
… first two OK
… third is like ISO 9000 approach?

<alastairc> qv/

GV: provide your own standard and test it?

RMB: yes

GV: so they can set whatever bar they want?

RMB: could take plain language as an example

<Zakim> alastairc, you wanted to ask about "approved protocols", and comment about org-internal testing in "test cases".

GV: think it is an intriguing example to reach some areas we haven't yet reached

AC: could this be accessibility-related?
… for example, someone sets own test to fail missing landmarks
… would that be in scope?

RMB: think it would
… WCAG could require landmarks
… but not the specific landmarks for the particular site
… the site would define that for itself
… and test that it meets it consistently

AC: what about external protocols like plain language?

<Zakim> bruce_bailey, you wanted to ask if protocols subgroup has wiki or Google pages

RMB: there is still on-going discussion about that in the protocols group

<Wilco> https://github.com/w3c/silver/wiki/Protocols

RMB: but imaginable to use both

KH: brilliant!
… seems like a great way for moving forward
… for example, for language-specific aspects

<alastairc> draft RESOLUTION: Group agrees to progress in this direction, working on the next steps

<SuzanneTaylor> +1 to Katie's enthusiastic reaction

WF: how much of this has been incubated
… how quickly can we expect this to become mature

<Jem> so it is like quantative vs qualitative...

RMB: some checks have judgement calls, like logical sequence
… on incubation, we could take years on this
… if we agree to move forward, could explore in just a few months
… can drop if not useful in a month or two
… meant to help us move forward, not be part of the document

AC: helping us categorize things

<alastairc> shadi: On a similar note, on the draft resolution, is it AGWG to progress on this vs sub-group to continue exploring?

<alastairc> ... similar to Wilco, is it exploratory and incubation, or the set direction?

<alastairc> draft RESOLUTION: Group agrees to explore this direction, working on the next steps from the presentation

<Ryladog> I think this is a set direction until or if we find it cant work. The proposal seems to help address some room for answers for places we have gotten stuck

RMB: need to bring back to the main group

<Ryladog> That matrix

<bruce_bailey> +1 on agree to try for sure

<Ryladog> +1 to MG

MG: way to help move things forward
… can work in small groups
… did not take long to try out 1.1.1
… bring back to see if it useful to people

AG: is the trial for 1.1.1 documented somewhere?
… would be good to see

RMB: have a rough draft
… want to clean it up then bring it back
… will try to have it by Friday

KH: seems quite clear to me
… can imagine the 1.1.1 example
… closer to what actually happens in the real world

<Zakim> bruce_bailey, you wanted to say i see in notes that subgroup is struggling with "protocol" term

AC: not every success criterion would have a space in each cell of the matrix

BB: seems the term "protocols" is ambiguous for many
… also history in how this term is used

<Chuck> +1

<Wilco> 0

<jaunita_george> +1

<Ryladog> +1

<Lauriat> +1

<bruce_bailey> +1

<MelanieP> +1

<SuzanneTaylor> +1

<ShawnT> +1

<Jennie> +1 (excited to try this!)

<GreggVan> +1

<ToddL> +1

AC: any concerns to moving forward with the draft resolution?

<laura> +1

+1

RESOLUTION: Group agrees to explore this direction, working on the next steps from the presentation

RMB: scoping sub-group meets every second Wednesday at 3pm ET

<Lauriat> +1 to Rachael on the "Aha!"

RMB: welcome others to join the call
… helped me understand the work

<Ryladog> PLease send the infor for the calls and groups

WCAG 2.2 Visual Controls https://www.w3.org/2002/09/wbs/35422/wcag22-visible-controls/

<alastairc> https://www.w3.org/2002/09/wbs/35422/wcag22-visible-controls/results

Examples provided are not enough #1443

<alastairc> https://github.com/w3c/wcag/issues/1443

<Chuck> I will back Jemma up

WCAG 2.2 Visual Controls

<alastairc> https://raw.githack.com/w3c/wcag/issue1443-visible-controls/understanding/22/visible-controls.html#visible-control-innerline-1

Examples provided are not enough #1443

alastairc fixed verb tense

alastirc: regarding adding success example, we can do that but that would not block this issue of adding failure example.

alastirc: any questions?

michael: these is effective example.

<alastairc> draft RESOLUTION: Accept PR 2200

<sarahhorton> +1

+1 ;-)

<GN015> +1

<ToddL> +1

<jaunita_george> +1

<Chuck> +1

<ShawnT> +1

<Francis_Storr> +1

RESOLUTION: Accept PR 2200

Explicitly say that this mechanism is visible #2072

https://github.com/w3c/wcag/issues/2072

After some discussion, Frances Storr proposed a response, basically: It doesn't seem to improve the SC, and doesn't match other uses of that term.

David agreed with it

gn015: love to cross check the conformance with other rules

alastairc: it is compatible other rules

like "mechanism available"
… it is used at audio control, etc...
… if you find any descrepancy or unclarity, let us know.

gn015: I will do cross check.

+1

<alastairc> draft RESOLUTION: Accept response to #2072

<Ryladog> +1

<Chuck> +1

<sarahhorton> +1

+1

<jaunita_george> +1

RESOLUTION: Accept response to #2072

Adobe Comment #1888

https://github.com/w3c/wcag/issues/1888

Rachael created PR 2234 to address it with an update to the understanding document and a new technique, along with a response in the thread.

https://github.com/w3c/wcag/pull/2234/files

<alastairc> In some cases, all items within a component of the page, such as a design canvas, are editable. Adding the information needed to identify the user interface components would reduce users' ability to understand and use the interface. In this case, adding text programatically and visually adjacent to the component that states that all content is editable could serve as the indicator.

editorial adjustment suggested by Sarah

Editorial:

- large design [FIX TYPO] canvans

- <li> states that all content is editable [CUT could serve as the indicator]. </li></p>

Check for consistent use of terms component, sub-component, control, especially for the main instruction, “Add text to a control stating all content is editable.”, which I think should be “Add text to a component stating all content is editable.”

alastair: I can do check component part by Sarah's comment after the meeting

awk: question about scope of component. first one is that design canvas is component/

<Ryladog> sub-items

then what is the component concept in the second and third?

<Ryladog> and primart item?

we are talking about the component here in "In some cases, all items within a component of the page, such as a design canvas, are editable. Adding the information needed to identify the user interface components would reduce users' ability to understand and use the interface. In this case, adding text programatically and visually adjacent to the component that states that all content is editable could serve as the indicator."

<alastairc> "In some cases, all items within a section of the page, such as a design canvas, are editable."

michael: how about changing the name of control?

michael: text in canvas? and also we can use "target size" concept here too.
… we can work on on this because this is the first draft

<Zakim> alastairc, you wanted to suggest it is area or section of content

<Ryladog> Primary larger item, like a calendar, and each of its sub-controls

AWK: using illustrator, we have multiple layers ..
… part of way to discover the layer is using visible control like toggle.
… my concern is that when there are multiple of controls to interact with thousands items
… user may ask which visual control the user should use to interact.
… I don't see the way to explain clearly about intentionally obsucured visual control

michael:
… guessing things in layer effectively create

the mechanism
… I agree that we need more investigation
… like the case of color palette

<Zakim> alastairc, you wanted to talk to what scenarios we're confident about

alastairc: changing layers has point.
… I don't think the design area is not the inital case for this SC
… but there was the case in game(?) which has multiple controls
… if we craft some exception
… for items within editable area something like that?

<Ryladog> Items contained in an editable area

alastairc: design palette is difficult case. we would like to cover simple cases first.

michael: in word doc, there are muliple interactions such as right click, hightlight... isolating the object concept may be helpful.
… does adobe has unlock/lock in global level, awk?

awk: I can find out.

alastairc: we may verify those interactions whether it is using hover or click so on...

michael: current language does not have
… current draft "When user interface components are invisible until hover or focus makes them visible, provide a visible indicator that the components are available, except when:"

alastairc: current one does not consider to cover "select"
… that may be the issue
… clickablity of design interface seems to be the topic.

sarah: the intention is actionable when there is no affordance, context which tell users that it is actionable. - self explanatory
… we may want to pursue the direction of "self explanatory"
… becuse that is the intention of this sc

alastairc: sc is hard to take consider the context..
… we struggle with that. in addition, we had opposition from COGA
… like the people with memory loss issue.

michael: agree with that this is visual affordance by Sarah
… missing part is UIC part...visual affordance may help

alastairc: that is different from visual indicator..

michael: visual indicator vs visual affordance...

alastairc: we can form some scoping or exception for the SC

michael: I have been trying to bring designers to AG discussion

<AWK> +1 We can talk Mike

michael: I would love to explore how we can tackle this without causing problems to designers.

alastairc: can you add suggestion to the survey, Awk?

sarahhorton: I like to caution making complex what is really is
… it is definitely worth to looking into the complex case but I would recommend to focus on the core intension of this SC.

alastairc: my concern is about testability
… to say it passes or not.

michael: not only adobe tool, jamboard, miro tool also has this complex features.
… intentional interaction is important
… although I am not sure what this affects COGA

sarahhorton: regarding Alastairc's point
… I don't want to lose sight for the original intention as well as lose opportunity to solve the problem people already have.

<kirkwood> Can we actually depend on hover?

alastairc: WCAG 2.x framework, both need to be met - testability(broad coverage?) and original intention

<alastairc> https://github.com/w3c/wcag/issues/1888

michael: discussion of selectabilty

kirkwood: I have concern about "hover" part
… kiosk in public place does not have hover capability.

kirkwood: I am referring to any touch screen scenario

alastairc: so this sc scope is when people depend on hover design

michael: it will be really good to understand initial intent by COGA.

<alastairc> Q for COGA: is this for a touch environment as well as a mouse env? I.e. is it a problem where you select something to get more controls?

sarahhorton: I worked on remix of this sc quite while back. it was about affordance is.
… point John K made is really telling the story
… make things clearly actionable

<kirkwood> +1 to Sarah

<alastairc> acl gn

gn: click case is the worst case such as clicking white space to see any actional items

michael: two things - 1. adding focus wording
… focus was added later to close a loophole

<AWK> Requiring that users click in a blank area to identify a component sounds like a general usability issue. I don't think we see that being done anywhere today.

michael: so changing focus is changing the scope.

2. regarding the clicking blank space by GN
… entire interaction point is not covered, we may need to rewrite this.

<kirkwood> MG i think your interpretation of what Gundula said

<michael> That's not a uic

<Zakim> michael, you wanted to say not a uic

michael: any canvas would not meet the defintion of UIC.
… because canvas include mutliple UIC.

kirkwood: how about the case of ...
… dot, dot, dot

alastairc: dot dot dot as control to take action.

<sarahhorton_> This is the working document from the original intent: https://docs.google.com/document/d/1v9VN9JN7fWIv1fIlBNXRhibMnRavn0M2Bx6AohtZ_jc/edit#

michael: this discussion seems to be how we contain the scope...

alastairc: summary

<kirkwood> for example edit, move, change type size type controls in a design canvas

1. AWK case - extra and complex controls does not show necessarily on hover.
… 2. COGA question

<AWK> In tools like PPT you can insert a text area that has no border and if you don't add text to it right away you might need to click around to find it again (or use select all to find all objects, which can be overwhelming).

-... isolating this sc from hover only is beneficial or not

alastairc: we will try to answer to above two questions

alastairc: awk and michale can do some follow up for the survey.

Summary of resolutions

  1. Group agrees to explore this direction, working on the next steps from the presentation
  2. Accept PR 2200
  3. Accept response to #2072
Minutes manually created (not a transcript), formatted by scribe.perl version 185 (Thu Dec 2 18:51:55 2021 UTC).

Diagnostics

Succeeded: s/https://www.invaluable.com/auction-lot/original-1920s-cappiello-king-liquor-poster-veuve-44-c-b7630c45a5/

Succeeded: s/to progress or to explore/AGWG to progress on this vs sub-group to continue exploring

Succeeded: s/are/is/

Succeeded: s/agree/agreed/

Succeeded: s/caes/case/

Succeeded: s/pallete/palette/

Succeeded: s/focus was added later to make this meaningful/focus was added later to close a loophole

Maybe present: AC, AG, alastair, alastirc, awk, BB, Editorial, gn, gn015, GV, KH, MG, RMB, sarah, WF