Silver Conformance Options Subgroup

27 May 2021


Azlan, JF, PeterKorn, sajkaj, sarahhorton, ToddLibby, Wilco

Meeting minutes

Agenda Review & Administrative Items

Janina: We have a draft report for May. Would like to propose we skip the third party contract, and come back after doing the rest.

Jeanne: Presenting on the 21st of June for AG.
… Does it need to be closer?

Janina: I think it's fine for this report.

Jeanne: Instead of monthly reports, have it be what we're proposing for 3rd party content.
… We should work on it in June, so we can get it approved in July

Janina: We can expect to do that. Should go back and conclude our principles.
… We did not get consensus, but got the general idea. There was a request to clean them up. We can do that in June.

Peter: Love the idea of delivering that. I'm concerned on how well we can reach consensus on proposed language. Maybe like last time, we report language to be discussed, as opposed to language that has the consensus.
… The way to deal with it may be outside conformance, but still live in some section of WCAG 3.

Janina: That is part of the April report. We have a more solid proposal in the May draft.

Draft May Report https://www.w3.org/WAI/GL/task-forces/silver/wiki/May_Report_to_the_Silver_TF

Janina: Lets start from the freely provided item.

Peter: Reading from use case C

Peter: I think this is a variant of something we discussed last time. Sites can do a fair amount to prompt third-party users, but it may not be feasible to insure they are without failure.

Janina: Most sites will probably not want to over-prompt, not to annoy contributors. Accessibility takes some skill, which is not going to be everybody.
… ATAG seemed appropriate. Areas of these websites are authoring tools.
… You want to generate accessible content with reasonable expectations. But you also want the authoring environment to be accessible.

Peter: We said we'd deliver our proposed language. The idea is some form of attestation. We have prompted users, and call out something is third party content which may not have the same level of accessibility.

Janina: That may be part of it, but what I'm wondering is if we can do more for content-free frameworks.
… Can we look at this mini authoring environment and might we be able to say more.

Peter: Coming to our journal site example, they are using UI components for third-party content.
… They use prompting techniques. Drupal has some some testing and validation that these techniques are used to address WCAG guidelines.

Jeanne: We intend to include ATAG in WCAG 3. If we said, for third-party content, that it needs to conform with the authoring tool related guidelines. We could say that if you meet ATAG 2.0 for your user generated content, you would pass.

<Zakim> jeanne, you wanted to talk about authoring tools and content free framework

Janina: Can you score that, or does it rely on attestation.

Peter: How does a third party validate? I guess just as normal. Whoever assess the site would look who they prompted.

Jeanne: Drupal has implemented ATAG 2. That is one of the features.
… In this case you can say, we're using Drupal, it meets ATAG 2, therefore we met the requirements.

Janina: That sounds like a kind of attestation. You're counting on their claim.

Peter: Just as you can validate that the alt tag exists, you can potentially validate that when you enter text you are prompted.
… What you may not be able to do is adjudicate how well the prompting works.
… It doesn't necessarily solve all our problems, but moves it further.

Janina: That is what I was trying to get it. I think with this kind of environment we'll end with content that is all over the map.

Peter: We'll have the separate challenge, that clear language and all other SCs have. Which is how much of it can you do automatic.

Janina: Familiar. We're reusing it in this context.

Wilco: Just wanted to voice my support for this. Agree with Peter.

Janina: Do we need to add to our May report?

Peter: Yes. To say we see this addressable by incorporating ATAG, prompting site visitors to ensure content is accessible.
… Then it is up to ATAG to decide what is the bronze / silver / gold bar.

Jeanne: The attractive thing about this is to solve the partial conformance thing of WCAG 3.
… An organisation can't say partial conformance, they'll be sued.
… We have a two-fold approach. To improve the third part content, and two is to solve the third-party content challenge.
… It's not perfect on either side, but it advances the bar.

Peter: One other thing was to identify where one can find the third-party content.
… So a visitor would have a way of knowing what part of a site may be of a different level of accessibility.

<sajkaj> https://docs.google.com/document/d/1GyUYTnZp0HIMdsKqCiISCSCvL0su692dnW34P81kbbw/

Peter: Third party content identification may be part of the solution.

Wilco: We should be careful with this that it is not fully accessible. It is a way to draw boundaries about what an organisation is responsible for, vs what their users are responsible for.

Peter: Reading use case B

<jeanne> ATAG 2.0 <- https://www.w3.org/TR/ATAG20/#contents

Peter: There are several things going on in this use case. We may not have accessibility supported methods.
… I think there are issues of accessibility supported, and then there are issues of timeliness. Is the workload something that an agency can deal with.

<Azlan> https://developer.mozilla.org/en-US/docs/Web/HTML/Element/s

Peter: Do we want to recognise that in WCAG 3?
… Those feel like the two issues raised by this use case.

Peter: The other piece of it is what do you do. If there is no technology accessibility supported. You can either stop using the non-supported technique, or we're going to say there is no supported technique, we'll move to it once it starts to exist.

<ToddLibby> A little late, I had to look it up, but in CSS, "marker" could possibly be used for line numbering.

Wilco: Description tracks might be a better example of this. It is in the HTML spec, but it is not supported in any browser.

JF: We can't provide a description track in text, and have AT read it.

Janina: Via TTS. We wrote the spec, nobody implemented it.

Janina: So replace the use case with this. It is on the edge of what is possible.

Jeanne: The timing aspect of that use case, it keeps coming back up. We have a timing issue of accessibility.

Janina: It is the next item in the report. I suggest we move forward to that.
… I don't know if this is a principle we haven't yet written, but it is something we need to say in WCAG.

Peter: Reading Instant Accessibility Is Infeasible section

<JF> +1 to remove "snail mail" - we also state "physical mail" which is less pejorative.

Peter: I wonder if, trying to up-level the use case into a concept. The concept is that there are policies that are at odds with publishing accessible content in a timely fashion.
… Not entirely dissimilar from contracted third-party content. For example copyright may be at odds.

Janina: And there may be a firehose of content, where content can only be made accessible when there is a specific need.

Peter: This may also be a case of attestation. Some statement about during periods of high volume, we will fall below our SLA.

Janina: If it's not a brand new SLA, the content below its time frame, there will be a small percentage of content that is without its accessible alternatives?

Jeanne: Accessibility standards today are a snapshot in time. We keep talking about issues on the content over time. That is a pretty radical departure of what we have today.
… I think we need to call that out as the issue, in this idea of wanting more time to make it accessible.
… There is no mechanism for it today at all, and we don't have a historical example that is a reasonable solution.
… I think it's an issue that has to be clarified. Is there a mechanism that we can include? I don't know if there is.
… Aspirationally, it'd be cool to do this.

Sarah: One source of confusion may be the term content. We're focusing on content and not functionality. I thought we were talking about third-party components, as well as third-party content.

<JF> +1 to Sarah. Not all web *stuff* is content (component libraries are widgets, not content)

Janina: In my mind it has always been both.

Sarah: The examples under use case B and C are all content oriented. It may make sense to talk about those things separately. The requirements may be different.

Janina: There is the content that you read, but then the process of putting it there is different.

Sarah: Last week we were talking about add-on functional components that users interact with.

Janina: They could be?

Sarah: Things like on the Amazon website, the functionality on the site is developed / maintained by Amazon, but there is chat functionality that was built by a third party.

Whether Amazon is responsible of that.

Sarah: The third-party content is not interactive content. It is static.

Sarah: The term "content" is a bit confusing, and we'll want to make sure what we mean is clear.
… Maybe talking about captioning might be a nice use case. How much time after publishing a live video should an organisation publish the captions.

Peter: I like that idea. I suggest using audio description, just because we have good live captions, but we don't have mechanism for doing live AD yet.

Sarah: When we talked about third-party content last week, it seemed the use cases were more focused on static content.
… The system is one component, but the content is another.

JF: I would suggest that multimedia may not be right. There are multiple techniques, but one of the things that happened in the browsers. You can put those in band. It makes the example muddy. The techniques are so different.

Sarah: I'm not talking about functionality, it's related to the time. How much after releasing the video do you have to have captions.

Janina: We'll have to say something.

Peter: Reading use case E

Minutes manually created (not a transcript), formatted by scribe.perl version 136 (Thu May 27 13:50:24 2021 UTC).


Succeeded: s/Janina/Sarah

Maybe present: Janina, Jeanne, Peter, Sarah