15:55:47 RRSAgent has joined #ag 15:55:52 logging to https://www.w3.org/2026/02/17-ag-irc 15:56:47 agenda? 15:57:10 agenda+ WCAG 2.x updates https://lists.w3.org/Archives/Public/w3c-wai-gl/2026JanMar/0070.html 15:57:40 agenda+ Review WCAG 3 Work Plan https://docs.google.com/document/d/1igBeBBh3ZWk1czXRO-4jhtSKyQ8Bc_-ZUoAalUeV-cg/edit?usp=sharing 15:57:46 present+ 15:58:15 Patrick_H_Lauke has joined #ag 15:58:23 agenda+ Work through an example ACT rule https://docs.google.com/document/d/1i_O5J7D4BWMZAB7NaD4iplDRuqOjCvCGAaHdAirLUhE/edit?tab=t.bo2e33svugxv#heading=h.8k3m19m21u2 15:58:23 present+ 15:58:36 present+ 15:58:53 bbailey has joined #ag 15:59:03 present+ 15:59:13 zakim, start meeting 15:59:13 RRSAgent, make logs Public 15:59:14 Meeting: AGWG Teleconference 15:59:27 GreggVan has joined #ag 16:00:08 present+ 16:00:31 Helen has joined #ag 16:00:34 shadi has joined #ag 16:00:37 janina has joined #ag 16:00:41 present+ 16:00:42 present+ 16:00:55 giacomo-petri has joined #ag 16:00:57 kirkwood has joined #ag 16:01:05 GN015 has joined #ag 16:01:10 present_ 16:01:11 Jennie_Delisi has joined #ag 16:01:12 present+ 16:01:13 present+ 16:01:13 present+ 16:01:13 present+ 16:01:16 present+ 16:01:17 present+ 16:01:17 LoriO has joined #ag 16:01:20 present+ 16:01:22 present+ 16:01:30 present+ 16:01:37 CClaire has joined #ag 16:01:40 ShawnT has joined #ag 16:01:44 present+ 16:01:45 present+ 16:01:47 present+ 16:01:49 present+ 16:01:59 AlinaV has joined #ag 16:02:08 scribe+ 16:02:12 present+ 16:02:16 stevekerr has joined #ag 16:02:23 present+ 16:02:32 LenB has joined #ag 16:02:34 laura has joined #ag 16:02:36 present+ 16:02:45 tayef has joined #ag 16:02:47 present+ Laura_Carlson 16:02:48 Rachael: Asks anyone new that wants to introduce themselves. 16:02:54 Adam_Page has joined #ag 16:03:02 present+ 16:03:03 present+ 16:03:11 present+ 16:03:13 Helen: From the ACT group, and a liaison for writing rules. 16:03:26 AWK has joined #ag 16:03:27 https://w3c.github.io/wcag3/guidelines/ 16:03:33 present+ 16:03:41 julierawe has joined #ag 16:03:44 jtoles has joined #ag 16:03:49 present+ 16:03:50 jkatherman has joined #ag 16:03:54 present+ 16:03:56 Jon_Avila has joined #ag 16:04:00 present+ 16:04:07 present+ 16:04:10 listserv email on pre-cfc: https://lists.w3.org/Archives/Public/w3c-wai-gl/2026JanMar/0073.html 16:04:54 Thank you to editors for already incorporating my feedback from last week! 16:05:06 Rachael: Announcements. In pre-CFC point on the next update to WCAG3 guidelines. Asks for review on the pre-CFC content to ensure there's anything major that might be missed. Reminder: going to be going into 'developing' stage, and they are looking for discrepancies and things that are missing. Next would be to get public feedback on the different 16:05:06 provisions and incorporating it. 16:05:11 zakim, take up item 1 16:05:11 agendum 1 -- WCAG 2.x updates https://lists.w3.org/Archives/Public/w3c-wai-gl/2026JanMar/0070.html -- taken up [from Rachael] 16:05:31 mfairchild_ has joined #ag 16:05:32 Rachael: Agenda item is WCAG 2 updates. Asks Patrick to review it. 16:05:38 Makoto_U has joined #ag 16:05:48 present+ 16:06:21 Patrick_H_Lauke: One week into the review. Open to comments and asks for thumbs up, or provide feedback if there's something missed, or problems that may have not been considered. One more week to go. 16:06:36 zakim, take up next item 16:06:36 agendum 2 -- Review WCAG 3 Work Plan https://docs.google.com/document/d/1igBeBBh3ZWk1czXRO-4jhtSKyQ8Bc_-ZUoAalUeV-cg/edit?usp=sharing -- taken up [from Rachael] 16:06:40 Rachael: Process for WCAG 2 is a pattern that will be emulated in the future. 16:07:04 Jen_G has joined #ag 16:07:29 Wilco has joined #ag 16:07:43 Present+ 16:08:09 Rachael: Agenda item for WCAG 3 work plan using the process document. This is in a public folder, but not being called out. Manages process and ensures things are getting done. A high level overview of the document. 16:08:29 Francis_Storr has joined #ag 16:08:32 present+ 16:09:06 joryc has joined #ag 16:09:13 present+ 16:09:18 Detlev has joined #ag 16:09:24 present+ 16:10:49 Atya has joined #ag 16:11:04 Glenda has joined #ag 16:11:04 ... The first tab is provision management, goes through the steps to complete each provision. Provisions map to WCAG 2 are prioritized. Issues gathered by provision short name in the GitHub repository. Ongoing process to address the feedback. Going to work on the HTML methods and do an example as a group today. Revise the provision language if 16:11:04 needed, then send out a survey, address the comments, and escalate to the main group if needed. This cycle is expected to be gone through several times. Next revisit the guideline text and survey the guideline text. 16:11:50 present+ 16:12:00 SydneyColeman has joined #ag 16:12:09 present+ 16:13:02 Rachael: First tab also includes a process to Closing Issues. Next to each provision name, there's a link to GitHub with all linked issues to the provision. Reviewed the remaining steps under the Closing issues heading in the document. 16:13:07 q? 16:15:31 https://docs.google.com/document/d/1igBeBBh3ZWk1czXRO-4jhtSKyQ8Bc_-ZUoAalUeV-cg/edit?tab=t.0#heading=h.fe0ypliwvzkz 16:17:18 I believe that the conformance subgroup was initial time bound through start of March. Is this now extended? 16:17:28 Rachael: On the 2026 tab, are the work plan for the calendar year. For each tab, the document lists each calendar years' objectives, conformance, policy document, plan. The plan targets 8 provisions to survey. Goal is to get examples, think about it, and ask for input to start the iteration cycle. Wanted to be cautious of the timing of the Policy 16:17:28 Subgroup so the interested parties had time to contribute. 16:17:57 Rachael: On 2027 tab, Objectives section has started. 16:18:09 Jennie_Delisi - we'll extend as needed, but we will need to report back soon, and pick a direction (or 2) to work on. 16:18:38 Rachael: On the Definition of Done tab, ask for ideas and contributions for the content. Currently, has criteria for provisions and guidelines definition of done. 16:18:43 Thank you alastairc - important to share with group in case they had only blocked calendars for the original number of meetings. 16:19:07 q+ 16:19:14 Rachael: Future Work 2028-2030 tab contents has just begun. If you have changes, a concern, please feel free to email the chairs. 16:19:50 q+ 16:20:10 Wilco: Document seems 'barebones' to me. Looking for a timeline to get this to an actual recommendation. Also, need clarity on which provisions are included, and what are we expecting for the conformance model? 16:20:56 ack Wilco 16:21:12 ack GreggVan 16:21:20 Rachael: Decision points are in the document; as for the conformance model, that decision needs to be made int he next 6 months. And we need to revisit the timeline in the subgroup as a whole as to where we're trying to get to. Coverage-wise, the minimum is WCAG 2. Reiterate that the document is a working document. 16:21:43 q+ 16:22:06 q+ on averaging out 16:22:31 It may be advisable to clearly note on the document that it “not a public document” and it may change as “working document” 16:22:56 q+ 16:23:30 GreggVan: In favor of having a process that says what we need to do before we release without strict timelines. Worries that if a timeline is defined, and that timeline isn't met, it would not have the due diligence that each subgroup is doing. Wants to ensure that the subgroups are able to do all of the activities in the process before moving to 16:23:30 the next phrase. 16:24:48 ack hdv 16:24:56 stevef has joined #ag 16:24:56 q+ 16:25:10 present+ 16:25:11 ack alastairc 16:25:11 alastairc, you wanted to comment on averaging out 16:25:12 hdv: Understands that even though the group hasn't done this specific work before, we've done very similar work before. We aren't inventing something new, there's lots of overlap in WCAG 2. The new things are the better conformance model, the structure is better, and making improvements to the provisions overall. Thinks its realistic to do the work 16:25:12 in the timeframe that's suggested, but is confident this group can pull it off. 16:25:32 +1 to hidde's optimism 16:25:46 q+ 16:25:55 alastairc: There's going to be an averaging out with the work throughout the year. There are some provisions that may take longer than others, but expects the work to average out month-to-month. 16:25:57 ack shadi 16:25:58 ack shadi 16:27:10 ack Wilco 16:27:11 shadi: From the description, if we're striving to have something comparable coverage to WCAG 2.2, what are the benefits or the improvements we are gaining? Also, thinks the plan is missing a narrative about how we are rewriting the guidance. 16:27:28 +1 to Shadi 16:28:02 q+ 16:28:02 Did we go massively over time for WCAG 2.1? 16:28:15 s/we are rewriting the guidance/we are addressing "complex challenges" listed in the current charter 16:28:34 Poornima has joined #ag 16:28:36 ack Rachael 16:28:45 Wilco: Curious about Hidda's comment. Doesn't know where the optimism comes from, as historically the group has gone massively over a deadline. We might end up in 2030 with something that may make the standard easier to read, and might improve things for people working in the field, but doesn't significantly change things for people with 16:28:45 disabilities. 16:29:04 q+ 16:30:17 q+ to ask if any known external date pressures (which we had for WCAG 2.1, 2.2, WCAG2ICT 1 and 2) at this time ? 16:30:21 s/Hidda/Hidde 16:30:48 graham has joined #ag 16:30:50 q+ to answer Bruce 16:30:52 present+ 16:32:04 q+ 16:32:07 ack GreggVan 16:32:09 Rachael: Trying to strike a balance. Asked for modular proposals and hadn't gotten one. Tried to start originally from the broad scope to understand what we needed to address in WCAG 3. At the same time, do our best to balance the schedule. Get through 200 provisions, learn the new process and work on the process to leverage all the existing ACT 16:32:09 rules, all the existing understanding documents and create provisions rapidly over the next few months. The Policy document covers topics that aren't conformance questions. The goal isn't only to have WCAG 2.2 and then go beyond it. Wants to make sure there's time today to go through ACT rules. 16:32:29 q- 16:32:54 q+ 16:33:11 ack shadi 16:33:17 zakim, close queue 16:33:17 ok, Rachael, the speaker queue is closed 16:33:18 GreggVan: Comment was WCAG 2, 2.1, and 2.2. It was not massively overtime based on chairs expectations, rather, the group was forced to set timelines. Standards take time to author. 16:33:37 +1 Shadi 16:34:20 q- 16:34:49 +1 Shadi 16:36:33 ack bbailey 16:36:33 bbailey, you wanted to ask if any known external date pressures (which we had for WCAG 2.1, 2.2, WCAG2ICT 1 and 2) at this time ? 16:36:34 Definition the policy document would be helpful. 16:36:36 Where are the significant gaps? (for us to address) 16:36:44 shadi: Worries that the policy document has become a bucket for 'everything else' and is expected to solve all these issues. One of these could be guidance to policymakers. WCAG 2 does, in the conformance section, talk about conformance claims. Thinks we need to go through every challenge and think about what parts of the challenge can be 16:36:44 addressed. What can be addressed in an informative guidance that we provide? What is it that we can't do anything about as a technical organization. Maybe that's something for the policy makers. Glad that the conformance subgroup has been recently restarted. Thinks we need to restart other subgroups. Wonders if there are other challenges being 16:36:44 considered by the chairs and not communicated. Biggest fear that we don't have a plan here. 16:36:56 EN 301 549 put pressure on WCAG 2.1, 2.2, WCAG2ICT publications -- and work was timely and responsive 16:37:25 ack alastairc 16:37:25 alastairc, you wanted to answer Bruce 16:37:30 bbailey: plus one to Greg and Andrew talking about WCAG 2.1 and 2.2. There's a lot of time pressures, but wouldn't characterize those as being 'well over' there. Do we know of any time pressures at this time? 16:37:56 Rachael: Understanding is at least 3 years before we have any pressures coming back. Alastair? 16:38:02 q+ 16:38:11 -1 Alastair, I think a big part of it is a lack of discussion 16:38:59 you have to be in the discussions... 16:38:59 alastairc: 3 years is roughly what I was going to say. We looked at various tricky topics. It's more been a lack of a solution rather than lack of discussion. There's an order, a sequence, list of potential contents of policy document, and we move on from there. 16:39:12 Rachael: Calling out that Wilco is disagreeing, and feels there's a lack of discussion. 16:39:18 Solutions that go into conformance 16:39:30 The policy doc is another way that we haven't tried before 16:39:43 shadi: If we're saying we don't have solutions, then how is a policy document going to help? We've failed at delivering responses to what's in the current charter. 16:39:43 +1 shadi 16:40:12 should note to the success of our standards… they have been applied outside our remit.(even legally) even VR apps. and more. I don’t underestimate the reach of WCAG 2.1 16:40:31 Rachael: We've discussed it a number of times, and would characterize this in a different way. Converstaions on these topics are part of the policy document and can get pushed back if there's another solution. 16:41:06 +1 to kirkwood 16:41:06 zakim, take up next item 16:41:06 agendum 3 -- Work through an example ACT rule https://docs.google.com/document/d/1i_O5J7D4BWMZAB7NaD4iplDRuqOjCvCGAaHdAirLUhE/edit?tab=t.bo2e33svugxv#heading=h.8k3m19m21u2 -- 16:41:06 shadi: Between the charter and policy, we need to clarify which goes where. Feels like it goes back and forth. 16:41:09 ... taken up [from Rachael] 16:41:29 Rachael: Moving to next topic: Working through an ACT rule example. 16:41:56 persistent figure captions: https://docs.google.com/document/d/1i_O5J7D4BWMZAB7NaD4iplDRuqOjCvCGAaHdAirLUhE/edit?tab=t.ji9ubcoz0mhp#heading=h.yqlujc8a0hnu 16:41:58 Rachael: We had discussed going through persistent figure captions today. Link posted above. 16:42:35 Rachael: This provision falls under your subgroup. This time is meant to be an exercise with the subgroup participation to work the rule. 16:44:09 Rachael: Last week, we went through an exercise on how to write ACT rules. Today, we want to try it out as a group and then do this work in your subgroup. Asks for the subgroup that owns Persistent figure captions to actively participate in this. 16:44:39 Patrick_H_Lauke has joined #ag 16:44:42 Rachael: (Reads description of the provision). Suggests to start with examples. 16:44:54 q+ 16:45:13 Wilco: Agrees to start with examples. Need a general idea of what it is we want, and examples. Think about the rules. 16:45:28 Rachael: See if we can find examples, links to examples, etc. 16:45:59 Patrick_H_Lauke: Coming in this cold. Hasn't looked at this, but what struck was that the actual definition of figure caption, it says it's always visible. 16:46:10 Example, hopefully passing :-) https://www.w3.org/WAI/WCAG22/Understanding/non-text-contrast.html#figure-buttons-no-visual-indicator 16:46:47 I wouldn't call those figure captions 16:46:54 +1 16:46:56 Rachael: Found an example of secondstarfwh.org. The text on the image is visible and text below them. Cards have a title, and then flip with another description. Asks if these are figure captions. 16:46:58 ack Patrick_H_Lauke 16:46:59 +1 16:47:32 Rachael: Andrew says they are not figure captions, and Patrick agrees. Wilco says this is an example of inapplicable. 16:47:44 GreggVan: Asks why they are not considered figure captions? 16:47:50 +1 to patrick's observation that wcag 2 terms have “gremlins” ... 16:48:02 q+ 16:48:07 according to the definition, the examples shown are figure captions. Is there anything missing in the definition? Like, a figure caption describes the figure. 16:48:09 q+ 16:48:24 joryc has joined #ag 16:48:28 q+ 16:48:32 GreggVan: Would they have been considered figure captions if the text on the back of the card? 16:48:40 it’s persistant? 16:49:01 and as commented, the "and is always visible onthe page" is probably not appropriate for the definition 16:49:03 Those seem like figure captions then. 16:49:04 Rachael: Read the definition of a figure caption, which included 'always visible on the page.' 16:49:12 q+ to work from corrected term ? 16:49:12 ack LoriO 16:49:29 LoriO: Depends on how the the text is put onto the picture, the screen reader may not be able to read that. 16:49:32 Example we are discussing https://secondstarfwh.org/ images below Our Programs 16:49:33 q+ 16:49:35 q+ 16:49:40 q+ to say: Are we over-indexing on the term "Figure Caption" 16:49:44 a caption describes an image. these go beyond that 16:50:08 those are decorative images on cards, a caption decribes the content of the image in context 16:50:08 +1 to gregg 16:50:14 ack giacomo-petri 16:50:18 GreggVan: Discusses captions definition. 16:50:33 Patrick, I agree, yet the current definition doesn't include the point "describes the content of the image in context" 16:50:41 q+ 16:50:49 q- 16:50:54 q- 16:51:01 Agree with Giacomo 16:51:02 q+ on alt-text vs caption, and the defintion 16:51:07 +1 to giacomo-petri. The labels aren't really captions. 16:51:11 q+ 16:51:17 +1 giacomo-petri 16:51:17 +1 to Giacomo 16:51:18 giacomo-petri: Caption should describe the image, but not the links. By clicking the element, it will be displayed additional description of the image or the topic they are talking about. In my opinion captions should be concise to describe the image. It is an accessible name for Links, buttons, cards, whatever they are. 16:51:19 ack jkatherman 16:51:22 Captions aim to bridge the gap between visual content and surrounding text 16:51:33 +1 to decorative 16:51:50 +1 to jkatherman 16:51:51 ack bbailey 16:51:51 bbailey, you wanted to work from corrected term ? 16:51:53 q+ 16:52:02 jkatherman: Adds plus one; wanted to state that they are decorative images used to list the programs. 16:52:36 ack Helen 16:52:42 bbailey: We've got to fix the working definition; it's circular. for a future caption, they aren't figure captions except for the fact they are always persistent. 16:52:51 Suggestion: Explanation or comment that accompanies a work of visual media 16:52:56 +1 to this being a very productive way to work 16:53:10 +1 Thanks Helen 16:53:12 +5 to that 16:53:12 ack joryc 16:53:12 joryc, you wanted to say: Are we over-indexing on the term "Figure Caption" 16:53:24 Helen: What is being discussed is partly why the rules are really importing. You're noticing gaps in what you're writing because the definition isn't clear. It will infer that the logic of the rule needs to be revised. 16:53:35 Here's a page with two types of figure captions. The first has an "i"/information icon that shows the caption and others are the standard type: https://www.mfa.org/article/2026/desk-and-bookcase 16:53:39 q+ 16:54:13 ack alastairc 16:54:13 alastairc, you wanted to comment on alt-text vs caption, and the defintion 16:54:34 joryc: A challenge with this particular phraseology is it's implicit connect to the figure caption element. It doesn't provide this in any meaningful way, and it doesn't do anything other than add a FigCaption rules. Consider we're getting into an area that is redefining what a FigCpation is, in the context of HTML. Consider a different term. 16:55:00 Wondering if we might say that figcaptions can be hidden if there is a way to show the figcaption, but the image and the caption need to be both viewable at the same time? 16:55:04 +q to point out the definition seems to be inaccurate, and we might need another rule for other kinds of text 16:55:08 Standard definition: A figure caption is a concise, descriptive text placed directly below a figure (image, graph, chart, or map) that provides context, explains its significance, and allows it to be understood without reading the main text. It typically includes a figure number, a title, and details about the data, methods, or key takeaways. 16:55:44 ack GreggVan 16:55:46 alastaircThere's a danger of mixing up the sort of alternative with caption. We want to avoid this. Agrees with Giacomo's comments, and put in a suggestion. We haven't gotten to writing the text yet, we want to improve the definition. We need to be careful of the examples provided, and ensure they are 'good' examples. 16:55:55 Hiding images prevents people with other types of disabilities who benefit from the captions but don't use AT from accessing. E.g. cognitive, some with low vision 16:56:08 Sorry - image descriptions 16:56:27 +1 to AWK two good example 16:56:33 q+ 16:57:49 GreggVan: First, don't take the 'always visible out' otherwise you automatically bring tooltips and alt text into your definition. The problem was that we started off saying these are examples of figure captions. We were confusing something by the example; images and definitions, and the problems with the definitions. In this case, we are not 16:57:49 talking about a typical image, we are talking about a button that may go someplace. 16:57:55 ack Jon_Avila 16:58:09 alastairc: Anyone in queue that can speak to the original using it on this? 16:58:33 alastairc - is my comment at :55 possibly the original reason? 16:58:50 Jon_Avila: Often the figures are used in works because it is a reference to in text to the figure. You need the association between the surrounding text and the label of the image. For example, 16:59:25 ... refer to 'figure 1' - you need to know where 'figure 1 is.' 16:59:32 ack Wilco 16:59:45 joryc3 has joined #ag 17:00:14 Wilco: Process-wise, we need to move on from these examples. We found two interesting cases. Decorative images (do we want anything specifically for decorative images), and ask if these are in scope. 17:00:16 q+ to ask if ACT rules frequently use n/a examples ? 17:00:20 q+ to say definition of decorative iimages 17:00:22 ack GN 17:00:22 GN, you wanted to point out the definition seems to be inaccurate, and we might need another rule for other kinds of text 17:00:34 ... To keep moving, asks to explore what a failure looks like? 17:00:36 q? 17:00:37 q+ to add a failing example 17:00:37 Because these conversations can go on for HOURS, trust me :) 17:00:47 +1 to AC proposed revised definition: Explanation or comment that accompanies a work of visual media 17:01:13 qq+ 17:01:58 ack Rachael 17:01:58 Rachael, you wanted to react to GN 17:02:20 GN015: We see a few things. Feels that the original definition does not clearly state that a caption describes the image. It just talks about the 'text nearby.' Also, see in the examples where text is informative, though it is not describing the image, provides other information. The end user doesn't care about controls, they care about 17:02:21 association. 17:02:22 ack kirkwood 17:02:49 A figure caption is a concise, descriptive text placed directly below a figure (image, graph, chart, or map) that provides context, explains its significance, and allows it to be understood without reading the main text. It typically includes a figure number, a title, and details about the data, methods, or key takeaways. 17:03:01 kirkwood: Bristles at the former publishing term and the one that is standard was posted in IRC above. 17:03:26 alastairc: Requirement about having figure caption is that it is persistent. 17:03:35 placed directly below? 17:03:43 kirkwood: Confused about visually persistent - does that mean visually? 17:03:45 mfairchild_ has joined #ag 17:03:57 q? 17:04:13 alastairc: Reads definition. 17:04:16 ack bbailey 17:04:16 bbailey, you wanted to ask if ACT rules frequently use n/a examples ? 17:04:40 alastairc: Points out visually and hidden ones? 17:04:49 Yes, inapplicable examples 17:04:54 scribe+ 17:05:12 bbailey: Sometimes ACT rules have examples that is not what the rule is applying to. 17:05:20 q? 17:05:32 Yes 17:05:37 bbailey: Is it true that ACT rules often include/mention things that are not covered by the rule? 17:05:41 scribe- 17:05:53 Wilco: They all do, there are inapplicable examples. 17:05:54 q? 17:05:59 there needs to be a programmatic relationship between the caption container and the content it describes 17:06:02 ack GreggVan 17:06:02 GreggVan, you wanted to say definition of decorative iimages 17:06:35 Here's another example of an initially-hidden caption: https://www.mfa.org/exhibition/the-bold-and-the-beautiful (see the "i" button on the image in the lower left) 17:06:47 q+ to ask about scenarios and breaking things down 17:07:13 GreggVan: We also need to add to the definition that sets the tone. 17:07:14 I agree, most images are not decorative. 17:07:37 +1 to Gregg regarding decorative. 17:07:37 GreggVan: We're leaving too much out of decorative, and leaving it as something people can put on any picture. 17:07:53 In publishing we would say they (decorative) add to “the voice” of the writing. which a PWD would be entitled to 17:08:04 q+ to also provide decorative example 17:09:17 q+ to say a failing example would be very helpful 17:09:24 GreggVan: We should say that every picture should have a caption and the definition of that is something that is always visible. I think we should stick to what a caption is except it has to be above/below. We shouldn't try to force a rule into a definition. 17:09:30 That's not what inapplicable means. Something is inapplicable if the rule shouldn't be used. A rule about buttons should not be applied to hidden buttons for example. 17:09:35 q+ 17:10:07 GreggVan: If you have a pass/fail NA, if this rule does not apply in this circumstance, how do you test this? The rule should say what tests to use in this situation. Treating the ACT like the normative testing for the guidelines, but I don't see a 1:1 relationship between them. 17:10:11 ack giacomo-petri 17:10:11 giacomo-petri, you wanted to add a failing example 17:10:17 agree NA doesn’t practically work. 17:10:26 q- 17:11:15 giacomo-petri: I acknowledge that we can refine and reduce the applicability of the rule based on our needs but if they think that if we want figure captions to be persistent, we need to cover similar scenarios as the one Rachael mentioned. 17:11:34 q+ 17:12:44 Rachael can we record that in the doc? 17:12:46 ack alastairc 17:12:46 alastairc, you wanted to ask about scenarios and breaking things down and to also provide decorative example 17:12:47 giacomo-petri: The user agent takes over and the author has no control over it. We need to understand how to manage scenarios where the caption is managed by the user and not the author. 17:14:22 +1 17:14:29 ack Francis_Storr 17:14:30 alastairc: My subgroup went through an example (scroll to the template). Each of these rules is likely to be a subset of the full requirements/guidelines. We were tackling flashing on the page less than the threshold frequency. 17:14:32 alastairc: The specific scenario is that the text is associated with an image but hidden by default, until you hover or focus on it. Question for the ACT folks— does that sound like a reasonable requirement? 17:15:00 https://html.spec.whatwg.org/multipage/grouping-content.html#the-figure-element 17:15:19 ack Wilco 17:15:20 Francis_Storr: Having a quick look at the text in the document and it's probably worth pointing out, figure as in HTML figure, images are probably the most commonly used types of figures. 17:15:39 It seems like there are at least 4 things - 1. figure missing caption, 2. caption not associated, 3. caption not persistent, 4 caption not sufficient to act as caption. 17:16:00 Wilco: Can we record some of these examples? Tooltips are good to consider to make it a failure; does it matter whether it's above or below an image? 17:16:07 q+ on not-applicable, helpful for scoping, but what about when the requirement is asking for absence of something? 17:16:23 +1 to Wilco that "tool tip as figure capture" has good potential for failure example 17:16:43 Wilco: Wondering if there is a scenario where captions might go away after a certain time? e.g. 10 seconds after it appears? Would that be considered a failure example? Is it inapplicable? 17:17:02 tooltip is a failure if it cannot be navigated to, displayed using keyboard (or equivalent) 17:17:17 Wilco: If you have an image and click it, and a modal shows up with the c aption — you'd want the modal to be persistent but the image doesn't have a caption — is this an inapplicable example? 17:17:54 why would a caption that is displayed/dismissed on keystroke be a failure? 17:18:01 Wilco: We need a discussion on whether the requirements shoudl actually fail or allow tooltips because they're edge cases — assuming we know the requirement that it implies that. 17:18:06 ack alastairc 17:18:06 alastairc, you wanted to comment on not-applicable, helpful for scoping, but what about when the requirement is asking for absence of something? 17:18:30 q+ 17:18:41 alastairc: Gregg had a comment around NA (not applicable) — NA is demonstrating when something is not in scope, not that the testing is not working. Because of our definitions and scope of requirement, these are not applicable. 17:19:00 alastairc: Question for the ACT folks — what happens when the requirement is asking for an absence of something? 17:19:21 ack GreggVan 17:19:22 alastairc: In some scenarios, is it okay to ignore NA? 17:19:24 q+ 17:19:27 q+ responding to Alastair 17:19:28 I agree that no flashing is a pass. 17:19:58 +1 to Gregg's comment that if a provision is not applicable, it should 'pass' 17:19:59 LoC examples of initially-hidden figure captions hallway down page: https://www.loc.gov/#reuse 17:20:17 GreggVan: You have a pass/fail/NA — if the test is NA, it should pass. The question is then what does NA mean? If it is not referring to the clause in the requirement about why it is required then it is not an NA, it passes. 17:21:27 if NA is an answer the rule is broken. it should be rewritten 17:21:36 GreggVan: If you have a requirement that is true and there are 5 ACT rules that relate to that and all of them have to be true in order for it to pass, andi n other cases it's OR— one of them needs to be true in order to pass. These shouldn't be 5 ACT rules— should be one ACT rule that says it must do this and so on. 17:22:27 q? 17:22:35 q+ 17:22:40 q- res 17:23:00 GreggVan: I think it's great to have them but we need to clarify whether or not passing them is required or not required, or sufficient/helpful. The ACT are techniques, whether it is sufficient by itself or was contributory and each of those had a test procedure associated with it. I am confused about the role of ACT, and what passing them meant. 17:23:30 ack Helen 17:23:35 GreggVan: When I hit the NAs, if it's not applicable — what am I supposed to do? If the NA is exactly the same as a requirement then it shouldn't be an NA 17:24:55 if is potentially NA then it’s not a requirement. It should be rewritten. 17:25:14 kirkwood - I think you're misundstanding how these are used. 17:25:25 q+ 17:25:28 q+ 17:25:33 Helen: It depends on how important the rule is. In applicability, you need to have it be clearly defined at the start so that people are aware of how it should work. These are part of the SC, and it's more about the fact if the testing criteria is applicable, it won't say you're passing, but will give the idea of how to successfully test this success criterion. 17:25:43 +1 that having N/A being a “pass” for an ACT rule is at tension with WCAG2 where N/A is a pass against **SC** 17:26:04 q+ 17:26:04 Helen: It doesn't have to be automation rules, including manual testing, to help people understand the wording in the success criteria. 17:26:07 ack Wilco 17:26:10 The pass examples also need to state it's persistent - not just that the figure element is used. 17:26:30 q- 17:26:36 Wilco: All of this has been touched on by John Eaves (?) in his presentation last week. 17:26:47 qq+ 17:27:01 s/Eaves (?)/Yves 17:27:14 Wilco: To alastairc's question on what is inapplicable if the page doesn't flash — anything that is not HTML is going to be inapplicable in that scenario. ACT rules are specific to technologies. 17:27:18 so putting a caption in a disclosure makes it not persistent? 17:27:42 s/John /Jean 17:27:49 If there are no flashing objects on a page is it NA, or a pass? 17:27:50 ack Rachael 17:27:50 Rachael, you wanted to react to Wilco 17:27:53 So we have to create ACT rules for every media type. 17:28:09 Wilco: You may want to write a rule specifically for, example, flashing in an HTML page. 17:28:18 q? 17:28:20 ack alastairc 17:28:27 Rachael: Deep discussion going into critiquing ACT rules, would prefer to move that out of this meeting. 17:28:51 q+ 17:29:05 tag ? 17:29:23 alastairc: ACT rules have been applied to HTML, CSS, and ARIA so far but in response to Wilco, going back to the flashing example, not sure how to make that HTML specific— can use JS to hook into rules. Going back to kirkwood: if it's NA, it's not a requirement, it should be rewritten. 17:30:05 q- 17:30:08 thanks :) 17:30:47 alastairc: This exercise is good for thinking through examples that are passing, failing, or not applicable, and these are things you might think are applicable but technically not. We've got normative text, equivalency understanding document, and methods, and granular rules on how to test this thing. 17:30:53 q+ 17:31:00 alastairc: For ACT, we're going to try to fit in more than HTML because some things aren't HTML-specific 17:31:01 ack GreggVan 17:31:27 GreggVan: (ask for link to resources on how to fully understand ACT) 17:31:40 ack Wilco 17:31:48 LenB has joined #ag 17:31:50 Wilco: Every subgroup has an ACT liaison. 17:31:51 q+ to say: I think distinguishing between HTML and Javascript with a primary purpose of rendering UI (e.g. React JSX) is maybe not helpful for ACT 17:32:55 Rachael: We've talked about definition, examples, and say we've gone through everything that passes and fails, or is not applicable — what is the next thing to do for a subgroup? 17:33:10 Wilco: The next step is to look at the applicability, the expectation. 17:33:19 Presentation (PPT and video) https://drive.google.com/drive/folders/1eeR7jFDC74lDalqXMYcr4HoeXf9F2cW6 17:33:44 Wilco: Applicability can go something like "this rule applies to any… informative image element with a descriptive text positioned above or below it." 17:33:56 or canvas 17:33:57 q? 17:34:09 q+ 17:34:19 joryc - can we stick to the exercise please? 17:34:21 Wilco: What specifically do we consider an image? Descriptive text, we will probably need a definition for that. 17:34:53 Wilco: Ambiguity comes when people can have different interpretations, and makes testing difficult. 17:35:13 Helen: It's good to have a glossary rather than redefining what's already defined. 17:35:23 Wilco: ACT has a lot of definitions already written that can be borrowed from. 17:35:35 Wilco - is there a central place of the definitions you've been using? 17:35:43 Wilco: For the Expectation section, test targets are the things that are applicable. 17:36:11 q+ 17:36:39 Wilco: (more questions to think about when writing these rules) 17:36:50 Heather has joined #ag 17:37:23 Wilco: How do you make this rule atomic? You may want to say that those are two separate requirements that need to be two separate rules, or easy enough to explain in a single rule. 17:37:31 caption not part of accessible name, is accessible description... 17:38:07 ack joryc 17:38:07 joryc, you wanted to say: I think distinguishing between HTML and Javascript with a primary purpose of rendering UI (e.g. React JSX) is maybe not helpful for ACT 17:39:00 joryc: The ACT rules need to be mapped to specific technology, and we're concerned about writing rules that apply to multiple layers of the web, and I think separating the concerns seems unnatural. 17:39:06 I agree separating html, JavaScript, and CSS seems problematic. 17:39:15 joryc: Can we declare that ACT rules need to be technology-specific? 17:39:55 ACT rules are not tech agnostic, so not appropriate 17:40:08 ack giacomo-petri 17:40:20 alastairc: I was encouraging that, but I suspect there is nuance to it. e.g. if you consider it an HTML rule even if it is generated by React but is what the browser receives and interprets… we're gong to start writing rules for that and for native apps. Some things are non-tech specific. Please raise concerns to your ACT liaison. 17:41:07 The browser receives and interprets the HTML, CSS, and JS equally as a unit to render the user experience, both visually in pixels and the accessibility tree. That feels like the unit of test to me 17:41:10 giacomo-petri: We have to consider how we are refining the applicability of the requirement. 17:42:20 ack GreggVan 17:42:32 alastairc: Might want to look at what the rule description is, what is the rule, the boudnaries, and what you're trying to do — and you might want to separate that out. 17:43:25 GreggVan: Going back to defining an image, if an image is in a button, does it have to follow every one of those rules? If not, we should either change the definition of image or remove it and have a new definition for "standalone image" or "informational/mood-setting image", etc. 17:44:01 GreggVan: We need to be careful that we don't have definitions that mean something in our heads but apply to something much broader. 17:44:47 q+ 17:44:51 GreggVan: ACT — would be good if they were listed as tested techniques rather than testing rules, and could be techniques that are sufficient or partial. 17:45:08 ack giacomo-petri 17:45:13 alastairc: ACT rules framed in WCAG 3 is still on the table. 17:45:17 I agree the rules may not fully cover all aspects of a criterion. 17:45:34 +q 17:45:52 q+ 17:46:26 ack Helen 17:46:33 giacomo-petri: (Asking about the flipping cards from earlier) Not sure why it is passing, because the requirement says that the images must be persistent. 17:47:03 Helen: The ACT rules are not written to be comprehensive as one rule = one success criterion — one rule tests a scenario of that, so it is a technique and a way of documentation. 17:47:15 ack bbailey 17:47:52 q+ 17:48:09 s/the flipping cards from earlier/the passed example 2 in the doc 17:48:21 ack Rachael 17:48:21 alastairc: In your subgroups, everyone should have one or more provisions that you're working on — please think about them and how this can be applied. Some may be straightforward, some may not be. Practice using this template and that will be part of our next work. 17:48:51 Rachael: We're asking everyong to shepherd requirements—- take a requirement and be responsible for it. We need people to take up to 4. Work on that particular provision or requirement in between subgroup meetings. 17:49:42 Rachael: I would prep for my subgroup meeting by pulling as many examples as I can that would pass/fail/not applicable — prep the document before getting to your subgroup, for all of us to finish in a reasonable amount of time. 17:50:05 We've 40 people on the call, we should have 40 provisions for survey next month (and some this month). 17:50:08 q? 17:50:48 present+ 17:51:03 Rachael: I hope everyone will take one provision this week and work til it's done, then pick another provision. Just be thoughtful of moving things forward in a regimented way. 17:51:20 q+ to ask about 1st line of first tab of ACT Rules 17:51:26 ack bbailey 17:51:26 bbailey, you wanted to ask about 1st line of first tab of ACT Rules 17:51:28 central location for subroups adnd proviisions list? 17:51:48 https://github.com/w3c/wcag3/wiki/Guideline-List 17:52:09 bbailey: Tests for WCAG 3 are going to be written using the ACT rules format? I don't remember a group decision on that. I think it is very wise that we try and make it work. Feels very aspirational. 17:52:11 q+ 17:52:18 q+ 17:52:21 ack Rachael 17:52:40 q+ on non-HTML 17:52:56 q- 17:52:59 q+ to say ACT Rules Format doesn't limit to HTML 17:53:15 Rachael: We are using ACT rules format for tests in WCAG 3. We will have to modify in some places for clear language. There will be back and forth with ACT but they've built us a structure that we believe will work. 17:53:31 i am comfortable with that paraphrase, thank you 17:53:40 Rachael: In this draft we used the previous test format and it is not as helpful for working through these refinement processes as the ACT format. 17:53:44 ack GreggVan 17:54:34 q+ 17:54:49 q+ 17:54:55 GreggVan: Echo bbailey, I looked over the materials and asked a few ACT people and beginning to get a sense of it today — the rules are like techniques and they only test some things and those things that are easily/automatically tested. We're using the ACT format for testing our guidelines and they specifically say that the rules do not test the guidelines. 17:56:05 On q to say that ACT rules are flexible enough to include manual testing techniques. And the fact that ACT rules are currently focused on just HTML…just means we need to broaden this ACT model where we need it. Without an objective test methodology, I believe WCAG requirements are too mushy. 17:56:06 GreggVan: What is it that we need the testing methods to do? For example, they need to work across technology, all technologies. They need to figure out how to test things that are not automatically testable, and be able to test whole requirements. 17:56:43 GreggVan: It's not going to be something we can use unless it's testable according to ACT. ACT doesn't apply to guidelines or doesn't use them for those technologies. 17:56:43 ack dan 17:56:43 Daniel, you wanted to say ACT Rules Format doesn't limit to HTML 17:56:57 +1 to GreggVan's callout of the need for a statement which acknowledges when ACT Rules do and do not include a comprehensive set to test the entirety of the provision. 17:57:25 present+ 17:57:39 FYI on the WCAG2 usage https://www.w3.org/WAI/WCAG22/Understanding/understanding-act-rules.html 17:57:43 Daniel: ACT rules format doesn't limit the rules to HTML, just happens that's what the group has been doing all these years. We have extensive experience in writing HTML-based rules and there's also ARIA, but there's nothing in the format that limits that. In the current charter, using the format to write WCAG 3 tests would bring improvements needed for the formats into the table and useful for later. 17:57:46 +1 to feedback loop to ACT rules 17:57:53 zakim, close the q 17:57:53 I don't understand 'close the q', alastairc 17:58:13 ack Glenda 17:58:13 Could you re-post the URL of ACT rules for WCAG 3.0 (can't find it...)? 17:58:54 +1 to Glenda, it's about working through examples and making things unambigous, not tech 17:58:57 Glenda: I get to work with Wilco and he does often focus on what can be tested in HTML automatically but I know that ACT also has the flexibility to cover things that cannot be tested. ACT started with what was black and white — Wilco knew from the beginning that manual testing is important. We just need to take ACT to the next level. 17:58:58 ack Rachael 17:59:40 Rachael: No, current ACT rules do not cover everything, but we are expanding the ACT ruleset for WCAG 3 because we require a generic rule as part of our passing so new technologies can pick up WCAG 3. We will revise, add new rules, and like techniques, have compound rules, and so on. 17:59:46 We should be careful to to somehow encode or prioritize things that can be tested NOW with traditional, static analysis, versus what will soon be automateable. 18:00:02 Glenda has left #ag 18:00:12 present+ 18:00:23 present+ 18:00:23 zakim, end meeting 18:00:23 As of this point the attendees have been Heather, Patrick_H_Lauke, alastairc, bbailey, janina, shadi, hdv, Jennie_Delisi, Rachael, Helen, GreggVan, kirkwood, ShawnT, CClaire, 18:00:27 ... eloisa, giacomo-petri, AlinaV, LenB, Laura_Carlson, Adam_Page, stevekerr, tayef, AWK, julierawe, jtoles, Jon_Avila, jkatherman, Makoto_U, Jen_G, Francis_Storr, Wilco, Detlev, 18:00:27 ... Glenda, SydneyColeman, stevef, graham, Poornima 18:00:27 RRSAgent, please draft minutes v2 18:00:29 I have made the request to generate https://www.w3.org/2026/02/17-ag-minutes.html Zakim 18:00:36 I am happy to have been of service, alastairc; please remember to excuse RRSAgent. Goodbye 18:00:36 Zakim has left #ag 18:00:37 present+ 18:23:51 mfairchild_ has joined #ag 18:31:58 mfairchild_ has joined #ag 18:40:36 Glenda has joined #ag 20:15:22 daniel-mac has joined #ag 20:39:54 daniel-mac has joined #ag 20:54:26 daniel-mac has joined #ag 22:42:52 mfairchild_ has joined #ag 23:00:09 kirkwood has joined #ag