13:45:47 RRSAgent has joined #matf 13:45:52 logging to https://www.w3.org/2026/01/21-matf-irc 13:45:52 RRSAgent, make logs Public 13:45:53 please title this meeting ("meeting: ..."), JJ 13:45:54 Zakim, this is MATF 21 January 2026 13:45:54 got it, JJ 13:46:03 Meeting: MATF 21 January 2026 13:46:05 chair+ 13:46:19 agenda+ Project planning update 13:46:22 agenda+ Language of Parts voting 13:46:28 agenda+ User interface component voting 13:46:35 agenda+ User Agent definition 13:46:40 agenda+ Cross-platform Framework definition 13:46:44 agenda+ Not modified by the author / Provided by the user agent or platform software 13:46:49 agenda+ Underlying layer 14:00:55 tayef has joined #matf 14:01:23 Tanya has joined #matf 14:03:06 Detlev has joined #matf 14:03:16 RobW has joined #matf 14:03:20 present+ 14:03:21 present+ 14:03:39 present+ 14:04:52 scribe: Detlev 14:05:01 regrets+ TimGravemaker 14:05:07 regrets+ quintinb 14:05:08 pauljadam has joined #matf 14:05:11 regrest+ shoobe01 14:05:28 move to next agendum 14:05:28 agendum 1 -- Project planning update -- taken up [from JJ] 14:05:34 https://github.com/w3c/matf 14:06:11 JJ: Worked with Tanja on milestones for publication 14:06:25 ...20% complete 14:06:36 https://github.com/w3c/matf/milestone/4 = Second Draft Milestone 14:07:20 https://github.com/orgs/w3c/projects/147 = Project Overview 14:07:51 JJ: (lookin gat githob project view 14:08:05 ...any questions? 14:09:03 JJ: Check at what issues you are assigned to - or look at the to-do list and assign yourselves 14:09:29 q+ tayef 14:09:31 ack tayef 14:09:31 present+ 14:10:09 tafef: Can you speak what you recommend on how to work within the group, what are the practices? 14:10:33 JJ: No strict processes - Steven had the same question 14:10:43 ...there is a slack channel 14:10:46 To join Slack: https://matf.to/slack 14:10:57 The channel is #matf 14:11:46 or work on GitHub issues, if you have questions or draft - or send an email to the group 14:12:09 JJ: Or reach out to JJ so he puts it in the agenda 14:12:43 JJ: Maybe we need some guidance how to work, that is still work in progress 14:15:25 move to next agendum 14:15:25 agendum 2 -- Language of Parts voting -- taken up [from JJ] 14:15:27 JJ: (explaining how to enter Slack channel 14:15:35 https://github.com/w3c/matf/pull/278 14:16:12 Navigate to Files Changed to review the changes 14:16:21 JJ: looking for feedback - if you have no feedback you can approve it 14:16:34 https://github.com/w3c/matf/pull/278/changes 14:17:32 JJ: (explaining to review or individually comment in Github) - you need to be added as reviewer 14:18:59 move to next agendum 14:18:59 agendum 3 -- User interface component voting -- taken up [from JJ] 14:19:04 https://github.com/w3c/matf/pull/280 14:19:26 https://github.com/w3c/matf/pull/280/changes 14:19:34 JJ_ check if you have comments to "user interface component"... 14:19:39 move to next agendum 14:19:39 agendum 3 was just opened, JJ 14:20:06 move to next agendum 14:20:06 agendum 4 -- User Agent definition -- taken up [from JJ] 14:20:13 JJ: Definitions: user agent 14:20:14 https://github.com/w3c/matf/issues/63#issuecomment-3719286879 14:21:09 JJ: Last meeting there was discussion, comparing to WCAG2ICT - does it need to include mobile apps UA? 14:22:24 Tanya: Analysed what is stated in WCAG2ICT and EN 301 549 and WCAG - like the idea to compare multiple definitions - conclusion is that the notion of UA is not applicable to mobile apps 14:22:45 ...EN uses layers of responsibility 14:23:27 ...hardware layer outside scope; non-web software has platform software, an umbrella term similar to UA but broader 14:24:15 ...reason was likely because UA was introduced for the web context and linked to UAAG referenced in WCAG 2.1 and 2.2 14:24:46 ...in mobile apps it does not fall under the definition of UA 14:26:01 Detlev: so mobile OS falls under platform software 14:26:34 Detlev: How do you apply this to an app that has web content? Would the app be a user agent for that web content? 14:27:06 Detlev: Does not webview content be presented by the App like a UA? 14:27:20 Tanya: No this would be part of the platform software 14:27:52 q+ 14:28:04 ack JJ 14:28:08 Tanya: Following the WCAG2ICT webviews should be treated as platform software 14:29:24 Tanya: If be follow the layers approach, you have platform software (iOS / Android) and the application, the the accessibility services provided by the platform which we cannot influence 14:30:58 Tanya: definition modified / not modified by the author, there are elements that are not influenced by the author if yoused unchanged and would then fall under the platform software layer; if not (i.e. has been changed) it is treated as part of application and be author responsibility 14:31:41 JJ: recap of layers - does the author have control over it, (e.g. gesture recognition) 14:32:11 ...platform software: see if elements are rendered by platform software and therefore exempt 14:32:15 q+ 14:32:45 JJ: Problem in guidance is the mix of native and mobile, web included in native apps 14:33:04 q+ 14:33:12 ...last layer application software running on mobile OS 14:33:21 q- 14:34:15 JJ: discussion whether rendering of webviews could be seen as UA inside native app 14:35:17 ...biggest change if that latest EN has clarification of web views 14:35:36 q? 14:35:42 Tanya: EN said when in doubt, refer to clause 11 14:36:26 q+ 14:37:00 User interface rendering research: https://github.com/w3c/matf/issues/84#issuecomment-3748794971 14:37:24 q+ 14:37:27 Detlev: raising the question how easy it is to determine of content has been changed by the author or is rendered by the platform layerr 14:37:52 JJ: Has a proposal 14:38:20 JJ: uta the link above 14:38:57 JJ: trying to disambiguate between native and authored renderings 14:39:47 JJ: cross-platform frameworks may lead to native renderings, but might not 14:40:27 JJ: there is a table in that rendering research issue 84 14:41:37 JJ: with Android, only Android views are part of the platform but in prarctice Jetpack compose would also result in native rendering 14:42:00 I can't comment on that research from this machine, but I'm not 100% sure that SwiftUI Implementation is correct. Semantics = Jetpack Compose. SwiftUI = AccessibilityNode? 14:42:14 JJ: not putting all the detail JJ explians here into the minutes :) - look at the issue 14:46:47 q? 14:46:54 ack Detlev 14:47:01 JJ: It is still difficult to determine id things are rendered natively 14:49:01 ack Tanya 14:49:04 Detlev: if you need to ask developers what framework of design environment they used they will expct you to give more specific dev advice 14:49:21 move to next agendum 14:49:21 I see a speaker queue remaining and respectfully decline to close this agendum, JJ 14:50:06 Tanya: We have the same problem - it is impossible to determine if native or authored - but the comment will tell that things might be exempt 14:50:52 ...our goal is to make apps as accessible as possible - if devs change the color of a toggle or come back and argue for an exception 14:50:54 q+ 14:51:14 Tanya: The user agent/platform exception was meant for behavior/colors that cannot be modified by developers 14:51:52 Tanya: The list JJ has worked out, only frameworks that are part of the platform should be included, making the list shorter 14:52:30 Tanya: Suggests to exclude "Material Components" from "native" classification 14:53:08 I've filed a Feedback Assistant issue with Apple for their Wheel Style Pickers not having sufficient text contrast for non-selected options. But that is text contrast and it does not have the user agent exception. Only Apple can fix the issue either way. 14:53:22 ...if we look at the UAs with their own guidelines applying to browsers, but for native this will be provided by the platform software, so feedback should go there 14:53:47 ack RobW 14:54:40 RobW: appreciates Tanya's point on UA and platform that may include a UA. 14:55:40 ...wonder if out definition excuses things, let people o fthe hook that would be covered by UA 14:56:33 ...as to how do you know if it is rendered by platform: example of including rich text in a webview - where does this fall? 14:57:20 JJ: Difficult to determine. We have the accessibility inspector but sometimes it is a bit flakey 14:57:46 ...what tools can we expect an auditor to use? 14:58:26 ...for app, you may also expect auditors to have mire tools than just the app on the mobile phone 14:59:41 ack Detlev 14:59:54 Zakim, close the queue 14:59:54 ok, JJ, the speaker queue is closed 15:00:50 ...the exception is there for situations where developers have no way to change things - problem that android elements it may be impossible to mark things as buttons 15:00:54 if a developer changes the color of their switches by making a custom switch then when apple changes the style of the switch, then their custom switch will look old and not match the new UI design, the happened with Liquid Glass 15:01:53 JJ: kets have the same items on the agenda next week - please make drafts for the definitions if you can 15:02:22 Zakim, list participants 15:02:22 As of this point the attendees have been Detlev, RobW, tayef, Tanya 15:02:37 present+ pauljadam 15:02:48 rrsagent, make minutes 15:02:50 I have made the request to generate https://www.w3.org/2026/01/21-matf-minutes.html JJ 15:12:46 rrsagent, bye 15:12:46 I see no action items