14:29:04 RRSAgent has joined #ag 14:29:09 logging to https://www.w3.org/2026/08/04-ag-irc 14:29:09 RRSAgent, make logs Public 14:29:10 Meeting: AGWG Teleconference 14:29:25 agenda+ Introductions 14:29:34 agenda+ announcements 14:32:00 MartijnHoutepen has joined #ag 14:33:49 agenda+ WCAG 2 proposed changes https://lists.w3.org/Archives/Public/w3c-wai-gl/2026JulSep/0075.html 14:34:10 agenda+ Paths report-back https://docs.google.com/presentation/d/1pApedUR1EFcQJKlcmnysoKpVGLbAhjivyTh9yXgfuLI/edit?usp=sharing 14:34:19 Bernardo has joined #ag 14:34:24 agenda+ WCAG 3 document breakdown https://www.w3.org/wbs/35422/wcag3-document-breakdown/ 14:34:43 agenda+ Can flashing requirements account for zoom? https://github.com/w3c/wcag3/issues/631 14:43:56 Zam has joined #ag 14:45:43 agenda? 14:45:49 present+ 14:51:02 Dirk has joined #ag 14:56:56 LoriO has joined #ag 14:57:07 present+ 14:58:00 HaTheo has joined #AG 14:58:10 LoriO has joined #ag 14:58:59 present+ 14:59:19 present+ 14:59:21 present+ 14:59:28 present+ 14:59:32 Heather has joined #ag 14:59:36 present+ 14:59:38 wendyreid has joined #ag 14:59:40 present+ 14:59:57 Charles has joined #ag 15:00:09 present+ 15:00:09 ashleyfirth has joined #ag 15:00:28 JamesH has joined #ag 15:00:34 present+ 15:00:48 present+ 15:00:53 present+ 15:00:59 CClaire has joined #ag 15:01:02 GreggVan has joined #ag 15:01:07 present+ 15:01:10 erinevans has joined #ag 15:01:14 present+ 15:01:16 Helen has joined #ag 15:01:17 present+ 15:01:23 present+ 15:01:31 present+ 15:01:38 giacomo-petri has joined #ag 15:01:39 present+ 15:01:42 present+ 15:02:11 Patrick_H_Lauke has joined #ag 15:02:21 present+ 15:02:34 scribe+ 15:02:40 BrianE has joined #ag 15:02:45 anton has joined #ag 15:02:55 AWK has joined #ag 15:02:58 Jennie_Delisi has joined #ag 15:03:05 present+ 15:03:12 zakim, take up item 1 15:03:12 agendum 1 -- Introductions -- taken up [from Rachael] 15:03:40 q+ 15:03:44 ack JamesH 15:03:45 bbailey has joined #ag 15:03:53 present+ 15:03:55 Jon_Avila has joined #ag 15:04:11 JamesH: James Harrington. Back with the group. Accessibility professional for 10 years, currently works for CVS Health. 15:04:20 zakim, take up next item 15:04:20 agendum 2 -- announcements -- taken up [from Rachael] 15:04:21 present+ 15:04:22 present+ 15:04:30 Matt_King has joined #ag 15:04:42 Rachael: TPAC is in October. Start making arrangements for attending. 15:04:47 zakim, take up next item 15:04:47 agendum 2 was just opened, Rachael 15:04:49 present+ 15:04:49 Patrick_H_Lauke: No announcements. 15:04:52 zakim, take up next item 15:04:52 agendum 2 was just opened, Rachael 15:04:58 zakim, take up item 3 15:04:58 agendum 3 -- WCAG 2 proposed changes https://lists.w3.org/Archives/Public/w3c-wai-gl/2026JulSep/0075.html -- taken up [from Rachael] 15:05:07 s/Patrick_H_Lauke: No announcements./Kevin: No announcements./ 15:05:20 Patrick_H_Lauke: Bi-weekly reminder about the email that went out last week with 4 items... 15:05:49 kirkwood has joined #AG 15:06:03 filippo-zorzi has joined #ag 15:06:19 On one particular item, GitHub collapses the long item, you may need to 'load more' on 4540. Link on the proposed response does not automatically expand. 15:06:21 graham has joined #ag 15:06:27 zakim, take up item 4 15:06:27 agendum 4 -- Paths report-back https://docs.google.com/presentation/d/1pApedUR1EFcQJKlcmnysoKpVGLbAhjivyTh9yXgfuLI/edit?usp=sharing -- taken up [from Rachael] 15:06:27 present+ 15:06:35 present+ 15:06:41 present+ 15:06:46 s/On/... On/ 15:06:51 Rachael: Path Subgroup is morphic. Reporting back to the larger group. 15:06:54 Slideset: https://docs.google.com/presentation/d/1pApedUR1EFcQJKlcmnysoKpVGLbAhjivyTh9yXgfuLI/edit?usp=sharing 15:07:03 ...Link to slide deck is above. 15:07:06 [slide 1] 15:07:12 julierawe has joined #ag 15:07:12 Laura_Carlson has joined #ag 15:07:14 present+ 15:07:19 rrsagent, make minutes 15:07:20 I have made the request to generate https://www.w3.org/2026/08/04-ag-minutes.html Adam 15:07:31 present+ Laura_Carlson 15:07:45 [slide 2] 15:07:46 ...will go through slides first, then invite members of subgroup to weigh in on slides, then open up queue from the larger group. 15:07:49 Makoto_U has joined #ag 15:07:59 present+ 15:08:04 ...Some definitions from WCAG included in deck. 15:08:34 Gez has joined #AG 15:08:39 jkatherman has joined #ag 15:08:43 present+ 15:08:50 present+ 15:08:53 ...terms are for context; we will survey these definitions at a future time. 15:08:58 [slide 3] 15:09:08 Patrick_H_Lauke has left #ag 15:09:17 Rachael: [reading slide 3] 15:09:22 shadi has joined #ag 15:09:56 [slide 4] 15:10:06 ...[reading slide 4] 15:10:07 RRSAgent, pointer? 15:10:07 See https://www.w3.org/2026/08/04-ag-irc#T15-10-07 15:10:22 thelounge has joined #ag 15:10:26 rayianna has joined #ag 15:10:35 present+ 15:10:44 [slide 5] 15:10:45 present+ 15:10:47 ...Outstanding questions noted, but not for today's conversaion 15:10:52 present+ 15:10:52 q? 15:11:05 ...stop sharing, inviting members of subgroup to chime in to correct or add anything that was just read. 15:11:09 q+ to ask 2 specific questions 15:11:10 LiaGoogle has joined #ag 15:11:13 present+ 15:11:23 ack Charles 15:11:23 Charles, you wanted to ask 2 specific questions 15:11:24 present+ 15:11:31 ...over to the broader group for questions, clarifications, concerns, etc. 15:11:38 q+ about process vs. path 15:11:44 q+ 15:12:01 +1 to charles - this is a problem with the definitions for these that should change. 15:12:01 Charles: On the first slide, you had definitions for interactive and non-interactive... 15:12:02 q+ to mention that inputs group wrestling with non-active interactive elements 15:12:15 Rachael: Definitions are out of scope for today's question, but has been noted. 15:12:20 q+ 15:12:37 q+ 15:12:42 ack GreggVan 15:12:59 Charles: On scope, one thing was could interfere with the path Is the thing that could interfere with the path inclusive? 15:13:13 LenB has joined #ag 15:13:15 Rachael: Going with linear queuing to invite others to answer. 15:13:23 present+ 15:13:28 q? 15:13:38 q? 15:13:40 ack bbailey 15:13:40 bbailey, you wanted to mention that inputs group wrestling with non-active interactive elements 15:13:52 GreggVanStatus is separate from identity, when you interact with non-interactive be in a separate group had come up with elements that are not interactive elements. 15:13:53 ack Matt_King 15:14:06 s/GreggVanStatus/GreggVan: Status/ 15:14:06 q+ 15:14:08 Rachael: Adding that note in the slides, but don't want to go in that conversation today. 15:14:30 qq 15:14:31 bbailey: The inputs group is wrestling this to the unavailable interactive elements to the ground. 15:14:32 qq+ Jon_Avila 15:14:48 q- 15:14:55 Rachael: We look forward to you bringing that definition back to us. 15:15:11 ack Jon_Avila 15:15:11 Jon_Avila, you wanted to react to Matt_King 15:15:15 thank you Matt_King for clarifying intersecting paths 15:15:18 ack Jon_Avila 15:15:18 q- Jon_Avila 15:15:53 Matt_King: About Interfering piece. As soon as you include a particular portion of the UI in a path that no other path could also include that same portion because the same element may be involved in completing multiple activities. 15:16:05 Jon_Avila: Please remind me of the process and path relationship 15:16:20 Rachael: Both terms are separate. Process includes every Paget that goes from end to end. 15:16:21 q+ 15:16:48 ...path is focused on the elements and can also span domains. There are slight differences in how they are used and how they are scoped. 15:17:17 Jon_Avila: Follow up. When that's included, what about error states and how that relates to the path? 15:17:45 ack wen 15:17:54 Stephanie has joined #ag 15:18:01 Matt_King: If you define the path for the beginning at a certain point on the page, it will assume that the point to navigate to that point is included in a separate path. 15:18:01 present+ 15:18:11 q+ 15:18:30 wendyreid: There will be cases to have a very strict path, and also may have slightly more expansive. 15:18:50 ... This flexibility allows people to create paths and testing units that reflect how their products are constructed. 15:18:53 q+ 15:18:59 ack GreggVan 15:19:02 ...It's going to be particular to your needs and org/team setup. 15:19:25 GreggVan: I think this is the best treatment of paths that I've seen. It's easy to confuse conformance. 15:19:43 ...What if a path starts int he middle of a page, but you can't to the middle of the page. 15:20:06 ...a path is just a report, if you got to this point and you started here and did this process or path, it would be accessible. 15:20:19 q+ 15:20:37 ...If it were for conformance, then you'd have to add things like that. But this is just for reporting. 15:21:04 ....If we get to the point of using paths, we have to include whatever you are doing in public reporting. You also need to include what is not included in the report. 15:21:15 ack Jon_Avila 15:21:29 Jon_Avila: Following up to Matt's question. 15:21:56 ...Would there be a prerequisite for a path? Or a definition of the relationships between paths? 15:21:59 q+ 15:22:20 ack awk 15:22:30 ...To Gregg's point, since it's starting from an element, is someone supposed to help them get to that point to start their experience if they can't get there? 15:22:42 q+ to answer AWK 15:22:55 q+ 15:22:57 q+ 15:23:07 ack Matt_King 15:23:07 AWK: Asking if the working group made a decision about whether paths were part of conformance or not, and believes the answer was 'no.' 15:23:26 q+ 15:23:28 ...Asking that question again, it's worth bringing up to the group. 15:23:47 Matt_King: I can try to address JOhn's question. This was intentional on my part; not sure about the subgroup. 15:23:58 s/answer was 'no.'/answer was 'not yet.' 15:24:20 ...No explicit statement on that you need to specify whether or not there are other paths that are a prerequisite paths or dependency paths. 15:24:32 ....But generally going to be the case that there's a lot of dependency paths. 15:24:42 ack Rachael 15:24:42 Rachael, you wanted to answer AWK 15:25:04 ...If you're evaluating a specific scope, there will be dependencies of any particular UI scope on any other UI scope. 15:25:15 q+ 15:25:21 +1 AWK and Rachael 15:25:30 ack GreggVan 15:25:37 Rachael: Chair hat on. I appreciate AWK question, and agrees that the decision has not been made. 15:25:56 ...I appreciate you highlighting that it is part of the question that's coming back to the group in the proposal. 15:26:15 s/JOhn/John/ 15:26:20 ...now would be a good time to consider whether it should be or not, and if you disagree with it or not. 15:26:32 ...this is part of the question we are asking here. 15:26:55 Matt_King_ has joined #ag 15:26:56 GreggVan: This has not been part of conformance. It can be in a statement of non-conformance. 15:27:06 q? 15:27:09 ack wendyreid 15:27:13 q+ to ask if we are straw polling for paths in conformance 15:27:22 ...I don't think we should ever endorse the idea that people with disabilities only have a right to access certain aspects, while we make other aspects available to everyone else. 15:27:25 +1 to Gregg 15:27:52 wendyreid: We should also fill people in with some more of the background on the path discussion. It's a big thing. 15:27:54 Jen_G has joined #ag 15:28:18 ...It's a concept that anyone who's done auditing on a product will be familiar with. 15:28:57 ...When communicating internally versus externally, the message may be different. 15:29:24 ...For example, when there are things like shared components, you may call out when these things happen. 15:29:29 ack AWK 15:29:31 q+ 15:29:37 Present+ 15:30:03 ...This humanizes what can be a fairly dense set of informational levels and it allows us to communicate it in a way which makes it generally understandable to all types of players. 15:30:13 AWK: I think we should be thinking about paths. 15:30:32 ....We have a problem with WCAG 2 today, there are very few sites that can conform. 15:30:47 ...We don't want to discriminate against people with different types of disabilities. 15:30:49 Wilco has joined #ag 15:31:04 q? 15:31:07 ...I'm not at all confident that we're going to wind up with a single level of conformance. 15:31:10 ack Matt_King 15:31:23 ...I think we need to be thinking about this as a component of conformance. 15:31:44 Matt_King_: Wrestling with the question about what it means to say that paths would/nt be part of conformance. 15:31:56 q+ to propose next steps 15:32:18 ...My understanding is when you make a conformance claim, you can make it for a page; it's not against measuring a path or measuring to a page. 15:32:19 s/discriminate against people with different types of disabilities./discriminate against people with different types of disabilities, but today sign language users can't access WCAG 2.2 AA content as one example. 15:32:36 q+ 15:32:40 ack charles 15:32:40 Charles, you wanted to ask if we are straw polling for paths in conformance 15:32:50 ...It's against the provisions. I thought all conformance claims had to be scoped in some meaningful way. So who cares what the scoping mechanism is. Does it matter? 15:33:13 Charles: Rachael, are we trying to answer the question of yay or nay, is this part of conformance? 15:33:36 q+ to give example where paths helpful to conformance 15:33:36 Rachael: Please weigh in on pros, cons and perspectives on this question. My suggestion is to weigh in on if it should be or not. 15:33:56 q+ to weigh in on considerations for conformance 15:34:16 ...Subgroups next step is to pull together the surrounding documents and informative content, and then circle back from that to understand what a conformance model might look like. 15:34:26 q- 15:34:31 ack Jon_Avila 15:34:38 ...Would like the group to express their thoughts on whether this should be in conformance or not, so that could be taken into account in the design of the proposal. 15:35:11 Jon_Avila: I think it hinges on a bullet on the updated path definition to where it says all elements and states within the scope of conformance claim must be covered by at least one path. 15:35:30 q+ to say what I think it means for this to be part of conformance 15:35:43 ...If you want something on the site/page to be covered in conformance, it would need to be covered in a path. 15:36:11 ...If we don't have paths as part of conformance, and we use paths for reporting and testing, then I don't think it makes sense to require a path for every piece. 15:36:23 ack ggreg 15:36:28 ack GreggVan 15:36:31 q+ 15:36:41 ...If that piece is already being covered by a non-path type coverage. I don't think we need to force people to create paths just for the purpose of creating paths. 15:37:17 GreggVan: There was a comment that everything needs to be scoped, so it doesn't matter what the scope is - we have units of evaluation for each provision. 15:38:06 q+ 15:38:07 ...If you defined conformance to all the paths, they would be good for assigning work and for reporting internally, and for reporting progress toward conformance. 15:38:36 ack bbailey 15:38:36 bbailey, you wanted to give example where paths helpful to conformance 15:38:36 paths could be helpful to encouraging conformance because it allows a claim on a smaller subset 15:38:38 ...asks for anyone who's arguing for path as conformance, to indicate how they justify that idea of having not everything accessible. 15:39:20 q+ 15:39:40 bbailey: I'm rebutting Gregg. I think having passes for conformance by having a smaller subset of things. I like the example of the footer. 15:39:51 ack charles 15:39:51 Charles, you wanted to weigh in on considerations for conformance 15:39:56 paths could be helpful to encouraging conformance because it allows a claim on a smaller subset 15:40:13 ...It would allow the site owner to make a conformance claim for shopping, but would include partially non-conforming pages. 15:40:36 Charles: I agree with Matt's description that a path is a method to define the scope of what the conformance claim is. 15:40:58 ...Heard the other concerns; if it's part of conformance, which I believe it can and should be, it still has caveats. 15:41:28 ...If you have a defined scope of the activity and getting to that activity, then we need to be able include that. 15:41:35 ack Rachael 15:41:35 Rachael, you wanted to say what I think it means for this to be part of conformance 15:41:58 ...Clearly identifying what's not in the path and what's not in the claim, I believe that is a good dependency to have if we put it in conformance. 15:42:17 Rachael: Chair hat on/off. Addressing Matt's question about what it means to be part of conformance. 15:42:47 ...We have conformance, we have the word term or phrase progress toward conformance, and we have partial conformance, which could be partial non-conformance. 15:42:52 ...These are all different. 15:42:56 s/I like the example of the footer/I like the example of the inaccessible footer on a path that is otherwise conforming/ 15:43:17 ... We we say conformance, a path would be sufficient when making a conformance claim. 15:43:31 q? 15:43:32 ...The concern that Jon raised is about the elements that are part of the path. 15:43:34 ack Matt_King 15:43:35 a 2.x way of making my point is that there is at least one non-interfering and non-disrupted path to point A of any scoped path. 15:43:45 q+ to discuss some needs which require a path 15:43:49 +1 for Rachael vision 15:43:50 ...By saying it's part of reporting and testing, I believe that we are saying that it is not part of conformance. 15:44:24 ...We are saying that you couldn't just claim conformance of a path. You'd have to include it in a set of pages, or a process, but you could reporting progress towards conformance against paths. 15:44:46 Conformance requirement requires complete processes. 15:45:05 ...My report that your process is conformant; you wouldn't be completing conformance against smaller units that span that space. 15:45:27 Matt_King: I think part of the problem is that I've never seen a conformance claim in over 30 years of working in the industry. 15:45:31 +1 to Matt_King. i have not seen a conformance claim since the days where people used the WCAG 1 gif in their footer circa ~2003 15:45:38 ack wendyreid 15:46:02 +1 to wendyreid 15:46:24 +1 to distinguishing between contexts 15:46:32 ...You could currently say that I want to make a conformance claim for a page, don't make paths for the sake of making paths. 15:46:43 ...I think every single element on a page is part of some type of user activity or flow. 15:47:06 ...Paths are really just a way to scope things by purpose, as opposed by their location. 15:47:21 q+ 15:47:30 wendyreid: Want to challenge Gregg's assertion that has automatically equal content not being covered. 15:48:04 q+ to ask for input from those who have not spoken and wrap topic 15:48:15 ...Condsidering the explanations of using paths in a conformance context, I think in reporting, this can be very infuriating because it depends on how the product team has chosen to segment their product 15:48:23 ack GreggVan 15:48:26 ...and the context of how they are explaining things. 15:48:48 ...I think this can be explained in informative guidance and we can provide people with ways to do that. 15:49:13 ...I think it's helpful to separate out those unites because they are both equally valid, but they are slightly different. 15:49:18 s/assertion that has automatically equal/assertion that paths automatically equals/ 15:49:48 q+ 15:49:50 GreggVan: Process is part of covering WCAG. If you have a login in order to shop, then your login page has to be included. 15:49:53 Sorry, need to drop. Will catch up with the minutes. 15:50:30 ack Jennie_Delisi 15:50:30 Jennie_Delisi, you wanted to discuss some needs which require a path 15:50:43 s/Sorry, need to drop. Will catch up with the minutes.// 15:51:14 ... It is critical to be able to include conformance claims so that enforceability in court based on what is being claimed. 15:51:54 ack Jon_Avila 15:51:55 +1 15:52:01 Jennie_Delisi: Are we comfortable saying it's not required? The challenge I'm having is that I can imagine that there could be a need for us in the future to say a path is critical. 15:52:35 Jon_Avila: Defining what is included in the path might be tricky, such as needing 'help.' What if someone needs the help link in the footer to complete the path? 15:52:46 ...It's hard for us to know when the user will and will not need that. 15:53:24 ...If you can define the paths and you can report on which paths are included, and then which part partially nonconformist or partially conformant. 15:53:41 ack Rachael 15:53:41 Rachael, you wanted to ask for input from those who have not spoken and wrap topic 15:53:56 ...Someone would then have to determine what is sufficient and if that is sufficient for whatever purpose. 15:54:21 Rachael: Chair hat off. Jennie, to your question - because we are proposing this as a unit of testing in addition to page views and elements... 15:54:30 would a format or channel for help ever occur in real-time within a session in order to complete an activity? 15:54:37 Paths give context. thereby opportunity to give support without need to have everything on a page. 15:54:47 q+ 15:54:47 ...I believe we would still have able to say, this particular requirement must be testing at the path level without this being considered a sufficient unit of conformance... 15:54:57 ...where a conformance claim could be made against only a path. 15:55:04 ack Matt_King 15:55:18 ...personally, don't a sufficient unit of conformance with addressing the concern you raised. 15:56:00 ...Chair hat on, I want to open the conversation to the concern raised, and encourage everyone in the group to weigh in. Very open to hearing everything; we are not voting today. 15:56:21 q+ to answer Matt's question to me -- or you can skip if you want to keep this tuine for others 15:56:21 Matt_King: I still have the same question over and over. Gregg, when you said someone goes into court and say, "I conform"... 15:56:32 ack kirkwood 15:56:48 ...it feels like very non-specific language. I'm really trying to figure it out. When we scope a conformance claim, you could have a list of page... 15:57:21 ...a list of views, and a list of activities. What are the purposes of those things? Still trying to wrap my head around that idea of having a unit of conformance that is comprehensive. 15:57:55 kirkwood: Want to mirror some of the things that were said around cognitive arena and emphasize having the ability to give information through a path... 15:58:19 ...I've been in the courtroom and making accessible paths either physically or digitally. 15:58:21 Matt_King i think conformance claims by URI are spelled out in https://www.w3.org/TR/WCAG22/#conformance-claims 15:58:42 ...I think it's a very important aspect and opportunity to make paths and making accessible paths that we can give support and information. 15:58:50 jkatherman has left #ag 15:59:39 ...it's important for us to move past this whole page view only approach. 15:59:47 FWIW, inputs working discussion page on active elements is at: https://docs.google.com/document/d/18RDQH1piGOCjae1RTymHaP-M_2RXHoXF0WDHqYvw14w/edit?pli=1&tab=t.qzkufnpvguun 16:00:29 s/page on active elements/page on interactive active elements/ 16:00:39 Rachael: Encourages all feedback to be heard. Please contribute here; if you email chairs then they will get it to the subgroup and make sure your voice is heard. 16:00:46 ...Time for a 10 minute break. 16:00:52 scribe- 16:01:05 rrsagent, draft minutes 16:01:06 I have made the request to generate https://www.w3.org/2026/08/04-ag-minutes.html bbailey 16:01:53 janina has joined #ag 16:01:57 present+ 16:02:24 5 minute break. 16:05:08 https://w3c.adam.page/agwg/timer/ 16:05:29 zakim, take up item 4 16:05:29 agendum 4 -- Paths report-back https://docs.google.com/presentation/d/1pApedUR1EFcQJKlcmnysoKpVGLbAhjivyTh9yXgfuLI/edit?usp=sharing -- taken up [from Rachael] 16:05:36 zakim, take up item 5 16:05:36 agendum 5 -- WCAG 3 document breakdown https://www.w3.org/wbs/35422/wcag3-document-breakdown/ -- taken up [from Rachael] 16:05:57 zakim, pick a scribe 16:05:57 Not knowing who is chairing or who scribed recently, I propose Rachael 16:06:02 zakim, pick a scribe 16:06:02 Not knowing who is chairing or who scribed recently, I propose Jennie_Delisi 16:06:10 zakim, pick a scribe 16:06:10 Not knowing who is chairing or who scribed recently, I propose Jennie_Delisi 16:06:12 zakim, pick a scribe 16:06:13 Not knowing who is chairing or who scribed recently, I propose ashleyfirth 16:06:22 zakim, pick a scribe 16:06:22 Not knowing who is chairing or who scribed recently, I propose CClaire 16:06:32 Sorry, can't do it today 16:06:34 zakim, pick a scribe 16:06:34 Not knowing who is chairing or who scribed recently, I propose BrianE 16:06:47 zakim, pick a scribe 16:06:47 Not knowing who is chairing or who scribed recently, I propose Stephanie 16:07:01 scribe+ 16:07:22 s/Sorry, can't do it today// 16:07:25 survey results: https://www.w3.org/wbs/35422/wcag3-document-breakdown/results/ 16:07:28 Rachael: Next thing is to pick up the WCAG 3 document breakdown 16:07:36 ... we've been discussing this for several weeks 16:07:56 ... we're starting at user-generated content, and only look at the survey for simplicity 16:08:03 ... [recounting results] 16:08:10 https://www.w3.org/wbs/35422/wcag3-document-breakdown/results/#xusergen 16:08:39 summary: https://docs.google.com/presentation/d/1l9VWE8GQNqJpPTNsfmPQudSUgQx-Oiis31N8OCD914k/edit?slide=id.g3f5cb4ac4bd_0_70#slide=id.g3f5cb4ac4bd_0_70 16:08:51 ... I put together a summary 16:10:02 ... from a majority standpoint, there's more towards including it in some way normatively 16:10:17 ... but can go both ways 16:10:19 q+ to ask about time from publishing user-generated content to evaluating it 16:10:47 ... going to the survey results 16:11:04 ... for user generated content, I'm going to talk to the callouts for those in favour of conformance 16:12:06 q+ 16:12:14 ... Matt King objects if not included, orgs should think about user-gen content on their platforms, we have the technology now 16:13:09 ... the group should not police third-party content, but we cannot control every user submission. The problem may not be the user but the technology or authoring experience 16:13:31 ... question of whether this is a legal issue 16:13:56 ... user generated content is a reality on many websites, but some of the requirements may be better covered by standards like ATAG 16:14:14 Q? 16:14:15 q+ to say that it is dependent on definition of UGC. I suggest -- "UGC: content added by users after the release of the product to users - where the author is 16:14:15 ... normative requirement could be merely enabling content to be conforming 16:14:26 Ack GreggVan 16:14:26 GreggVan, you wanted to answer Matt's question to me -- or you can skip if you want to keep this tuine for others and to say that it is dependent on definition of UGC. I suggest 16:14:27 GreggVan: I think it all depends on the definition of user-generated content 16:14:30 ... -- "UGC: content added by users after the release of the product to users - where the author is 16:14:59 present+ 16:15:02 ... if I'm paying users to generate content, that's not user-generated. If its an email program, I made it possible but users didn't, I'm not responsible for what they make 16:15:08 ... the definition needs to be clear about the boundary 16:15:34 ... especially around whether the user is paid or not 16:15:42 Ack Charles 16:15:42 Charles, you wanted to ask about time from publishing user-generated content to evaluating it 16:15:42 +1 to Gregg 16:16:08 Charles: My understanding of making a conformance claim is dependent on the fact that the claim is also bound to a period of time when I make the claim 16:16:17 ... rendered irrelevant as soon as changes are made 16:16:25 ... for UGC, all examples are time sensitive 16:16:26 q+ 16:16:45 ... would require a way to automate continuous evaluation of the contnet to even be able to consider it part of a claim 16:16:56 ... how would we identify a definition of UGC for conformance that is time-bound? 16:17:01 ack giacomo-petri 16:17:15 giacomo-petri: I think we are looking at UGC from the authoring tool POV, but not from the user generating the contnet 16:17:34 ... example, if there is an LMS that allows for the creation of accessible contnet, but the user generated inaccesible content 16:17:46 ... I want to make sure my tool makes accessible content, maybe even guides 16:17:56 ... I want to flag the authoring tool as conformant 16:18:02 ... but it's also the author's responsibility 16:18:27 ... the content generated may not be relevant, but must be evaluated, I think it's a matter of policies 16:18:28 an example of ugc might be adding a review to a ratings and review page for a product. as a user, that can be added, removed, and sometimes edited at any time. manual evaluation would have to be constant. 16:18:28 q+ to mention UGC challenges for U.S. govt with “FOIA reading rooms” and Regulation.gov where citizens can upload (for example) image-only PDFs. 16:18:33 Ack GreggVan 16:18:39 +1 16:19:17 GreggVan: Charles you raise an interesting point, if I release an email program, when I release it its accessible, people use it, I update, I make a new claim, I don't want it to inherit the content made in the meantime 16:20:09 ... to Giacomo, yes, we should change from author to "authorer", if someone uses a website builder like Wordpress, is the author Wordpress or the author of the website through wordpress 16:20:10 q+ 16:20:42 scribe+ 16:20:57 GreggVan: Wordpress can make a claim, someone who uses tools that can't create accessible content, B2B, B2C, the offering is what needs to be accessible 16:21:15 Ack bbailey 16:21:15 bbailey, you wanted to mention UGC challenges for U.S. govt with “FOIA reading rooms” and Regulation.gov where citizens can upload (for example) image-only PDFs. 16:21:21 UGC challenging for U.S. govt with “FOIA reading rooms” and Regulation.gov where citizens can upload (for example) image-only PDFs. 16:21:25 Rachael: We want to acknowledge ATAG is a separate group and we don't need to overlap there, just think of our scope 16:22:17 bbailey: Challenges with US Gov't, the idea of editing what citizens are submitting as public comment, you don't want to edit what they provide, there are approaches to how to handle some of this content 16:22:22 Q? 16:22:24 ... like running OCR on PDFs that are uploaded 16:22:33 Best we can do is ensure tools enable the creation of user-generated content. We need to define it for us. 16:22:36 Ack wendyreid 16:22:38 wendyreid: the tricky part is the overlap with ATAG 16:22:40 +1 to Bruce 16:22:54 ... because a lot of the boundaries of the authoring tool versus the actions of the author are defined there 16:23:02 ... it’s not helpful for us to re-define that in WCAG 3 16:23:04 Bruce could you put link in to your feference? 16:23:06 ... because it would add confusion 16:23:09 +1 to Bruce. Definite need, and need to consider efforts needed to retroactively make effective if you are the responsible "publisher" 16:23:17 rayianna has left #ag 16:23:19 ... but when we think about UGC, there is some challenge in how far to go 16:23:30 ... because technology has come along a way in terms of the interventions can take place 16:23:42 ... in the authoring experience itself, or even in post-processing 16:23:55 ... not all the time, but some of the time 16:24:04 ... we also have to consider authorial intent 16:24:15 ... there’s a reason why someone is making what appears to be an inaccessible choice 16:24:16 Q+ 16:24:41 ... they might be doing it with the intention of accessibility but in a way that doesn’t conform 16:24:49 ... so need to be careful about boundaries 16:24:59 ack Rachael 16:25:04 Rachael: At this point we are focused on whether we should have something in conformance that addresses UGC 16:25:07 q+ 16:25:27 ... I'm looking through the responses, one comment against conformance would be it should be handled by policy and procurement requirements 16:26:00 ... UGC should not form part of normative requirements, it's difficult to control content at scale, and verify 16:26:37 ... not convinced we should have it as an easy path, or possibility of conformance, 16:26:37 Regulations.gov accessibility statement (which, unfortunately, does not mention how they are running OCR upon upload): https://www.regulations.gov/accessibility 16:27:02 it’s a legal issue 16:27:10 ... deciding how to handle UGC is a legal, policy issue, not for us to decide, we can have the technical boundaries 16:27:18 s/feference/reference/ 16:27:46 ... clear boundaries of where responsibilities lie with content, and other standards like ATAG 16:27:58 ... any additional comments? 16:28:00 Q? 16:28:01 q+ to discuss need to know what people say 16:28:08 Ack GreggVan 16:28:16 could an example of ugc be a video of an event where the presence of flashing lights cannot be prevented or removed on a platform that automatically plays video? 16:28:35 q+ 16:28:52 GreggVan: I'm being convinced we should just have something in conformance that WCAG is a standard to determine whether or not something is accessible, but not where the responsibilities lie, for that we can refer them to the policy docs or applicable regulatory bodies 16:28:59 ... this includes UGC, archival information, etc. 16:29:16 ... anything else we've discussed, it really is a policy matter. UGC is one aspect 16:29:24 +1 to pointing to policy doc 16:29:31 ... regulators and enforcers need to keep all this in mind 16:29:33 +1 to Gregg refer peopel to the policy docs or applicable regulatory bodies 16:29:33 Ack Jennie_Delisi 16:29:33 Jennie_Delisi, you wanted to discuss need to know what people say 16:29:33 q+ 16:30:02 Jennie_Delisi: Going back to the comment on who is the author and who is responsible, when we have multiple parties responsible for the development of content 16:30:49 Ack kevin 16:30:56 ... there is a tool that is the original producer, then further tools that process or ingest that content, there's been situations where someone posts something inflammatory towards someone else, we want that person to be able to perceive that content and make a decision on it 16:31:29 kevin: Question, Shadi put me onto this, 5.4, what is the problem with what we have in WCAG 2 regarding this? 16:31:35 Ack giacomo-petri 16:31:38 + to kevin 16:31:39 Rachael: Any thoughts? 16:31:40 q+ 16:31:52 +1 to kevin 16:31:59 https://www.w3.org/TR/WCAG22/#conformance-partial 16:32:28 giacomo-petri: Last time I somehow introduced the idea of a conformance audit, this could be a place where we define the author generated vs user generated specifc to the audit being performed. In some cases we want to audit the content 16:32:37 Ack Wilco 16:32:44 ... but other times we don't want to include it, depends on who the author is 16:33:03 q+ 16:33:25 Wilco: I think 5.4 is really good, I think the thing that is missing is the additional requirement that you don't just say that this is UGC, but that this page conforms except this bit 16:33:30 q+ to reply to Kevin -- what else is wrong with 5.4 ? 16:33:44 +1 16:33:46 +1 to Wilco ... 5.4 + some ATAG-type requirements 16:33:47 to answer who is the most responsible for the accessibility of content? A: the stakeholder with the most $ £ € $ ¥ ;) 16:33:49 ... a 3.0 that does 5.4 plus allowing users generate content in a way that conforms to ATAG 16:33:59 Ack GreggVan 16:34:03 5.4 has a ‘within two business days’ parameter that points to the challenge of my previous question. some ugc is added hundreds of times per hour. 16:34:23 GreggVan: First, we need to change our language, partial conformance is not a good description of that content, there is no such thing as partial conformance 16:35:16 ... this paragraph is about something different, something conforms but came out of conformance due to UGC over time, but if you could prevent it vs not, it's different 16:35:18 W? 16:35:18 q+ 16:35:24 ... don't need to do things out of your control 16:35:26 Ack bbailey 16:35:26 bbailey, you wanted to reply to Kevin -- what else is wrong with 5.4 ? 16:35:52 bbailey: 5.4 has the wrong name, should be UGC not third-party content, since third-party content has come to mean something else that should be part of conformance 16:35:59 Ack kirkwood 16:36:05 +1 to Bruce 16:36:28 kirkwood: I agree with bbailey, this is really tough, we're not really defining the user here, but then saying "user generated" but we have different understandings of the user 16:36:30 q+ 16:36:42 ... had this issue with multiple layers of who is responsible 16:36:42 Perhaps content outside of author control and move this into policy recommendations 16:36:53 Ack giacomo-petri 16:37:19 giacomo-petri: To add to John's point, I've seen instances with third-party content with UGC, carousels of instagram posts for example 16:37:30 ... another layer of complexity 16:37:54 Rachael: Ok, I'm going to move on to other standards alongside WCAG, such as Design for All, how would we include it? 16:38:05 ... would it be in conformance, would we make it informative? 16:38:18 https://www.w3.org/wbs/35422/wcag3-document-breakdown/results/#xother 16:38:20 ... [reviews results] 16:38:21 q+ 16:38:39 ... just going to read different takes 16:38:48 ... [goes over comments] 16:43:50 Ack GreggVan 16:44:19 GreggVan: Formal note, if we want our standard to be cited by others, it has to follow certain standards, the things we cite need to follow similar standards 16:44:38 wcag 2.x § 5.1 already cites and requires an external standard – RFC2119 16:44:43 ... we should make sure whatever we cite was created with the same considerations we have 16:45:21 ... separately from saying we'll cite it normatively, we should be harmonized so that there is no conflict between them, so that following one doesn't invalidate the other 16:45:39 ... if we cite something with cost, we demand people buy the other standard to use WCAG 16:45:40 +1 to GreggVan 16:45:59 q+ 16:45:59 Q? 16:45:59 1+1 to GreggVan 16:46:03 Ack Wilco 16:46:38 Wilco: We can have something more abstract, citing any specific standard is not the way to go, but offering up options as ways to build out and demonstrate conformance 16:46:58 ... 9001 and Design for All both are examples of ways to build up a11y programs and seem appropriate to me 16:47:06 q+ 16:47:26 Q+ to ask about assertions 16:47:33 ... D4A, as Shadi mentions, is not fully formed yet, more experience needed, but this feels like an important step to take to advance, we don't need to make it required, but strongly encourage 16:47:44 ... award some benefit for using these methodologies 16:47:47 Ack shadi 16:48:04 shadi: I was wondering if there was a problem statement, what are we trying to solve by doing that 16:48:12 ... getting additional points for doing this 16:48:27 ... are we missing something in WCAG, normative or informative 16:48:37 ... what are we missing, trying to solve? 16:48:39 q+ 16:48:46 q+ to respond to Shadi 16:48:50 Ack Rachael 16:48:50 Rachael, you wanted to ask about assertions 16:48:55 Rachael: We did discuss this, and did a spike on this, speak to the question 16:49:02 ... is an assertion sufficient? 16:49:07 Ack wendyreid 16:50:52 Ack Wilco 16:50:53 Wilco, you wanted to respond to Shadi 16:51:00 q+ 16:51:04 apologies / late regrets. i have to drop early. 16:51:07 q+ to say that these are not about the content but the organization so I am not sure how that fits in with WCAG core 16:51:15 Charles has left #ag 16:51:19 Wilco: the short version is I think we're missing that people follow a process to catch, fix, or prevent issues 16:51:37 ... and we have no way to reward or identify that that is part of the conformance picture 16:51:44 Glenda has joined #ag 16:51:44 ... they have means to continually improve 16:51:45 s/apologies / late regrets. i have to drop early.// 16:51:58 Ack Matt_King 16:52:01 present+ 16:52:09 s|apologies / late regrets. i have to drop early.|| 16:52:28 Matt_King: So with WCAG being an output and technical standard, in looking at other similar standards, it's usually the other way around 16:52:47 q+ 16:53:09 ... a lot of areas, especially in medical and transport safety, there is a process standard that says "you must do x", and the technical standard says "this is the way you do this testing or measurement", the process standard refers to the output 16:53:18 ... we're almost trying to do the opposite 16:53:30 ... it goes against other regulated spaces, upside down 16:53:33 Ack kevin 16:53:33 kevin, you wanted to say that these are not about the content but the organization so I am not sure how that fits in with WCAG core 16:53:41 +1 to Matt_King 16:54:07 kevin: Building on that, I'm totally supportive of these kinds of standards, there's an element of this is that WCAG is a ruler to determine the accessibility of a piece of content 16:54:22 ... adding in things like this actually changes WCAG and how it works, and not to its benefit 16:54:30 q+ 16:54:36 Ack Wilco 16:54:37 +1 to kevin 16:54:40 ... it expands it in a way that may not be as effective 16:55:05 Wilco: We're doing this a bit through assertions, requirements that aren't testable, they are kind of there 16:55:33 ... I do very much acknowledge its an odd duck, things do go the other way, but WCAG is so dominant that if it doesn't say this, no one will take it seriously 16:55:40 q+ 16:55:40 Q+ to note the question of boundaries in assertions 16:55:42 ... even if unusual, it's important for us to do it 16:55:48 Ack Glenda 16:55:49 I think the difference with the assertions is they are (or should ideally be) specifically targeting the accessibility of specific aspects of content. For example, a policy for sign language for live events 16:56:09 Glenda: If we don't include somrthing like this in WCAG3, we're really just rewriting WCAG2 without much of an expansion 16:56:28 ... is this how other standards do it? No, but we can make progress in accessibility by having assertions 16:56:32 +1 to Wilco 16:56:38 +1 to Glenda 16:57:05 +1 to Glenda 16:57:07 ... people could lie, but let's not get caught up in that, it may not be objectively testable, but if a company says I have an accessibility policy, an SLA, a CAO, and my breech rate is only x, those are auditable 16:57:13 yes WCAG is a temperature test. Let us not break the thermometer. And continue to give tools and procedures for the doctor. 16:57:14 ... those will move the needle 16:57:21 Ack shadi 16:57:53 shadi: Thanks Glenda and Wilco, I'm hearing different things, individual point by point requirements, specific content tequirements, that makes sense and fits into what Kevin is saying 16:58:35 ... making a piece of content more accessible, but I don't think ISO 9001 or D4A would give you that, so the question was do we reference other standards in their completeness, I don't think that works, but if the problem we are trying to address 16:58:46 -1 to Glenda, I think WCAG 3 has more within it than only this and I am not sure my concern is about whether this is testable or not 16:58:50 Ack Rachael 16:58:50 Rachael, you wanted to note the question of boundaries in assertions 16:58:53 ... where indvidual assertions address them, different questions, worth considering 16:58:54 +1 to what Shadi is saying about including assertions in WCAG instead of referencing other standards 16:59:02 Rachael: We have two things going on here 16:59:21 ... we have some consensus around not having whole standards but do we have parts we want to capture 16:59:31 ... what are the boundaries of our assertions 16:59:50 ... organizational structures and policies, could be assertions and could be informative 17:00:13 ... chair hat off, personally, for me this feels like its own level, a level of organizational requirements and it resonates with me 17:00:15 1? 17:00:16 Q? 17:00:29 ... we'll pick this up next week, thanks all 17:00:40 ... we'll bring things back to the subgroup, have a good week! 17:00:50 Thanks everyone 17:01:09 zakim, generate minutes 17:01:09 I don't understand 'generate minutes', Rachael 17:01:10 janina has left #ag 17:01:14 zakim, make minutes 17:01:14 I don't understand 'make minutes', Rachael 17:01:19 zakim, end meeting 17:01:19 As of this point the attendees have been Rachael, LoriO, MartijnHoutepen, Bernardo, Adam, HaTheo, Heather, Zam, wendyreid, JamesH, ashleyfirth, kevin, Dirk, erinevans, Charles, 17:01:19 rrsagent, make minutes 17:01:20 I have made the request to generate https://www.w3.org/2026/08/04-ag-minutes.html Adam 17:01:26 ... Helen, CClaire, GreggVan, giacomo-petri, Patrick_H_Lauke, Jennie_Delisi, bbailey, anton, BrianE, Matt_King, filippo-zorzi, graham, kirkwood, julierawe, Laura_Carlson, Makoto_U, 17:01:26 ... Gez, jkatherman, rayianna, Francis, thelounge, LiaGoogle, LenB, Stephanie, Jen_G, janina, shadi, Glenda 17:01:26 RRSAgent, please draft minutes v2 17:01:27 I have made the request to generate https://www.w3.org/2026/08/04-ag-minutes.html Zakim 17:01:32 I am happy to have been of service, Rachael; please remember to excuse RRSAgent. Goodbye 17:01:32 Zakim has left #ag 17:04:20 HaTheo has joined #AG 17:07:05 HaTheo has joined #AG 17:12:14 HaTheo has joined #AG 18:44:33 Glenda has joined #ag 18:51:19 Glenda has left #ag 20:40:55 Glenda has joined #ag 20:41:18 GlendaSims has joined #ag 20:45:51 GlendaSims has joined #ag 20:59:14 GlendaSims4 has joined #ag 21:10:26 GlendaSims has joined #ag 21:11:41 GlendaSims has joined #ag 21:23:35 GlendaSims has joined #ag 21:30:15 GlendaSims has joined #ag 21:30:54 GlendaSims has joined #ag 21:31:44 GlendaSims has joined #ag 21:38:33 GlendaSims has joined #ag 22:16:42 kirkwood has joined #AG 23:32:43 kirkwood has joined #AG