Context: This meeting was held while w3.org was down including IRC, email, and some referenced content. We managed queue via zoom and took notes here. This was not an ideal solution and we appreciate our members and scribe volunteers for their patience with working through it.
WCAG 3 Project management and communicating with Chairs/Editors [5 minutes]
WCAG 3 Update from Test Reliability (Joint ACT/Silver) [10 minutes]
WCAG 3.0: Programmatic Language example of revised outcome and method
WCAG 3 Third Party mediahttps://www.w3.org/2002/09/wbs/35422/med_con/ [20 minutes]
WCAG 2.2 Focus appearance https://www.w3.org/2002/09/wbs/35422/wcag22-focus-appearance-enhanced2/
Nobody introduces themselves
No new topics
Extra comments Rachael
Last week we went through Code of Conduct
We send the link via mail
group-ag-chairs@w3.org (Chuck, Rachael, Michael, Alastair)
group-ag-plan@w3.org (chairs + Wilco, Jeanne, and Shawn)
Wilco joined as project manager
Action: Add emails and when to use them to AG page
WF: Did some experimenting with how ACT might fit Silver
WF: ACT doesn’t cover all possible SC
WF: how can we do outcomes and methods with consistent results
WF: (explaining what has been done with an Outcome and related tests / Methods)
MG: Love to identify the gaps, what can’t be automated (compared to the easy identifiable checks)
MG: automated + process
WF: ACT is not only about automated, but about what consistently can be tested
JF: I don’t see “how to test” section, programatically yes, manually not yet
WF: it will be up to the tester what toolset to use and how
Cybele: Is there a conflict between ACT documentation and Plain Language?
WF: We did not discuss that yet
Jeanne: there will be a “get started section” in the Method
JF: I think Chris was correct on methods with test, so how do we integrate within method tests , https://www.w3.org/WAI/GL/WCAG3/2020/methods/images-of-text/ where the test tab has automated and manual. So breaking that down to manual test and automated tests within the ACT framework.
Wilco: Tests will be implemented within CfC approved applicability and expectations. Work in process.
Mike: On building on Wilco was showcasing on lang, expectations org needs to meet. The degree of expectations to fill gaps through manual testing , etc. is the work that we need to do. Start what we can test for automated, identify gaps, then test for manual tests, etc.
Rachael: Link to survey is available https://www.w3.org/2002/09/wbs/35422/med_con/results , based on a single question , which was around is it better to acknowledge that some organizations can’t fix third party OR is it better to not acknowledge . There is conversation around topic and exceptions around third party ability to fix.
Peter: This is based on the second set of issues. We started out with third party media. We generalized on media considerations , i.e. aren’t limited to third party media. I agree with your proposed reframing.
Janina: I agree. We came down to three things. Do everything they can. Inform users what they are going to get. Allow users to raise appropriate issues within the entity / platform. Also talked to legacy rights around methods / techniques. Do what you can with what is within the platform's capability. Talks to example of gov’t agency and issue around handwritten letters in presidential library and inaccessible content as a legacy example. I.e what is reasonable vs. what is capable within the platform itself.Should have the ability to create content that is accessible. Comes down to the scalability issue on what makes the threshold. Also comes down to rights, for example captioning and ownership or version that has captioning. Could be more than one version per licensing. We may not have methods or techniques for everything. I.e. street views may be usable and accessible by some, but not all. We want to be helpful to move accessibility forward. Do what you can. Let users know what they are getting w/ current status of the platform. Allow for issues to be raised by users.
Chuck: Chat question on extending the survey over concerns over current format.
Rachael: I think using surveys as it to get through today, but using a basis for conversation today. No formal decision.
Jennie D. : Third party media concept of hand written comment. I.e. decision making and participating in gov’t. If people can’t participate due to not being able to view, what is the basis for this or will this be part of the paper?
Janina: Citizen vs. official ?
Jennie D. : Citizen.
Janina: Citizen would be user generated.
Rachael M: After AGWG call, question came up can revised media be included. Survey was sent out. I’m pausing as I’m requesting to either 1) extend survey or 2) a new survey Straw Poll: A) Extend Survey or B) No decision today, allow a new survey
Rachael: Extending the survey seems to be the way to go per zoom chat votes.
Peter: We’ve been discussing something that should be in methods. Once scoring is set, I think we will have a better process for media being used. We are trying to explain the floor and then providing content on top of that moving forward, based on importance.
JF: In Janina's discussion, she outlined the iterative approach of if we can’t do something now, but can do something in the future.(Potential Protocol) Gradual increase of accessibility over time is welcome. Don’t let perfect be the enemy of good.
Greg: The link is broken, correct?
Rachael : Yes. We are talking about the topic of the survey. Content to be reviewed of survey is down).
Cybele: Is it possible to create an alternative mechanism to Collect end-user based third party content use cases to draw wider participation within community groups and wider AGWG community; this could be an interesting idea worth exploring. Might support a shift in the understanding of the problem. Use cases may drive conversation and other gaps may not be totally understood from the end user perspective. Perhaps an alternative mechanism to collect that info. :)
Greg: We should be careful to not write forward thinking text. Speaks to “possible” vs. you must do phrasing. I.e. technology that can read handwriting may be future technology that archives can use to parse. Don’t require it be done , i.e. alt text MUST be used but the option is there for technology to be used to do so. On Peter’s note , on items out of control, user’s content vs. not user content generated, distinguishing those is important.
David MacD: On legal aspects of media alternatives , I would like to press the pause button on it for a couple of reasons. Do we have domain expertise? Referencing WCAG 1, we didn’t let JavaScript off the hook. As a result , ARIA was introduced. We want to challenge ourselves to improve. Precedence is important to review when making a decision on making decisions and sending signals of direction.
JF: On media, media is more than just videos. Talks to other types of media, we don’t want to narrow what media means. I.e.
Rachael:
Chuck : switched things around but didnt change what things mean
Rachael: Results of the survey were 1 keep 4 move focus-appearance
Alastair : Tricky bit is that we want to set up focus indicator so that it .. could change min are to contrasting area
Gower : Fading to white discussing the halo indicator and treatment that won’t meet minimal area of contrast, but crucial for
Second point: put minimal are first then talks about contrast then adjacent
More description in understanding *delete link*
David M: nervous that people reading for first time I’m concern that hidden is separate thing. Wondering about confusion with it.
Rewording movement creates a conditional statement
Gower: did tackle initially, when in separate bullet. Think it can be addressed in understanding document
Had slight concern about havin that bullet first. But confident it can be handled in understanding document
David M: just want to be clear in first line
Wilco: think entirely hidden at top puts too much focus on it… maybe suggest instead of focus indicator must meet 3 requirements, instead say AREA of focus indicator meets criteria
Have we tried that?
Gower: the two scenarios: 2 color indicator, can’t have meet both. Half will always fail.
… same thing with halo affect it makes almost impossible to do
Alastair: if siad minimum area of focus indicator meets the following criteria then can drop that part. The tricky bit is the adjacent contrast. Might make simpler making it adjacent to component rather than entirety of focus indicator. Might be easier.
David: wondering if item of focus not entirely hiddin. Maybe in prose after to talk about things not covered up. Reducing conflation cognitively
Rachael: like to focus on phrase “item covered” like wrap it up. David type in your proposal?
David: wouldn't be a bullet just prose underneath it
Rachael: option 1 and option 2 and option 3 typed in chat:
Option 1) Move the phrase "the item with focus is not entirely hidden by author-created content back to a bullet.
Option 2) Move the phrase to the end of the SC text
Option 3) Leave it where it is.
Wilco: we have several others including content focus that are in that pattern
Gregg: I would not want it to be a floating sentence
Rachael: type in option you wish in chat
Option 1) Move the phrase "the item with focus is not entirely hidden by author-created content back to a bullet. Option 2) Move the phrase to the end of the SC text Option 3) Leave it where it is.
Gregg: Suggesting to have two bullets at the
top level
Alair: reviews current option 2 edit, puts clause 3 items deep
Racheal reads current working wording…
When user interface components receive keyboard focus an area of the focus indicator meets the following:
Minimum area: The area is either:
at least as large as the area of a 1 CSS pixel thick perimeter of the unfocused component, or
at least as large as the area of a 4 CSS pixel thick line along the shortest side of a minimum bounding box of the unfocused component, and no part of the area is thinner than 2 CSS pixels.
Contrast: The area has a contrast ratio of at least 3:1 between its colors in the focused and unfocused states.
Adjacent contrast: Where the area is adjacent to the component, it has a contrast ratio of at least 3:1 against adjacent colors of the component or a thickness of at least 2 CSS pixels.
Also, the item with focus is not entirely hidden by author-created content.
Three different change sets, just working on ordering Q for now.
JF: AC has examples from current draft of Understanding doc, including visuals. We all seem to have similar understanding, but graphics seem like they could really help here.
Mike Gower: @wilco this seems to be a new pattern, no one used before in SC
MG. We eliminating detectable indicators with some of the proposals.
Rachael: We may be having multiple drafts here, so we may have to come back.
David M: Agree this is a new pattern, and very complex as compared to all other SC
GreggV: Complexity of two different colors needing have to have contrast, can imagine scenarios where the contrast disappears very quickly. Similarly, the one-pixel thick allowance might also be much less discernible than people think.
GreggV: I did go back and look at examples, but it was not always clear which were passing and which were fails.
AC: When lists are back up, I did send email reiterating some of the history
AC @ mike G to look at some of the examples
MG: we had two-tone example, need to double check that contrast is good
Rachael: We had another thread of questions regarding contrast and area.
AC: I have tried to incorporate Wilco’s suggestion from the survey. [see above]
Wilco: Agree, had similar experience with drafting, and have submitted PR.
MG: Whole series of focus examples. Fig 11 has focus indicator with gradient
… with Fig 12, example continues with indicator changing.
… consider inverse, with light part the inverted with dark, and I pretty sure might miss metrics. First time we have had this idea of area as opposed to just border.
AC: Agreed that this is difficult because sometimes contrasting area is sufficient, but not when we don't have spatial separation between border and backgrounds
GreggV: Is the idea being conveyed that one can just inverse and still have contrast?
AC: No, it was more to give an example with gradient, so requirement bullets could just focus on the solid color
Rachael: do we have more comments on these changes?
RESOLUTION: Accept the wording changes as documented in the minutes and incorporating Wilco's changes.
Wilco, wrt adjacent contrast, please take a look at my PR edit, as I think it is more explicit.
[double check on grammar / tense]
AC: If not adjacent, then focus is on change of contrast. That part can’t reference the color of the focus area, because too many colors to have high contrast with each other.
JF: Please add images to the spec for this.
DavidMcDonald looks at first few examples, see like “adjacent”
GreggV: Seems like “two color” example to me
Rochelle accepts