Meeting minutes
Agenda Review & Announcements
W3C Breakouts Day 2026
Submitted talks (the call ended yesterday): w3c/
<Fredrik> matatk: Lots of cognitive stuff on Breakouts Day. Lots of pretty interesting stuff. There is one about pointer vetns and gesture support.
<Fredrik> matatk: The pointer thingy is about making certain gestures first-class citizens of the web platform itself. Whcih has high a11y relevance, of course.
<Fredrik> matatk: There's also Web Sustainability stuff and some very intersting discussions about age-based restrictions on the Web.
<Fredrik> PaulG: I'm unfamiliar with the Breakouts day. What is required for gettign a topic and are there ones that APA want to get involved with or is this very informal everyone does what they want?
PaulG: Agent harnesses need to be closer to where devs live. UI systems are being made in low-level languages that supposedly work on all platforms so you have the same UI on a canvas on the web as you have on Windows terminals or on the Mac.
PaulG: Most or them are really in accessible.
PaulG: There's going to be a lot of people let behind on this issue.
matatk: All the inforamtion about Breakouts Day is linked on that GH repo.
matatk: Lowdown is being given.
DST Changes
US changed clocks, but other areas, such as Europe, haven't - so this meeting is an hour different for some (for a couple more weeks).
matatk: Well, it's that time of year. As usual. This madness is almost a month long for all of us. It is weird and not very wonderful. Keep a close eye on the W3C calendar.
Spec Reviews
MathML 4.0
request: w3c/
matatk: The new one!
matatk: Mathematics Markup Language. They had some big changes around 3.0.
CSS Gap Decorations
request: w3c/
tracking: w3c/
Eric_hind: I#ll have a look at it.
matatk: There's a lot of resources available on this one.
PaulG: The only gap we care about is when they delete scroll bars. Or something.
Fredrik: We'll tap Jason for the MathML.
CSS Anchor Positioning Level 1
<matatk> tracking: w3c/
matatk: Looking for people to edit that document.
matatk: We talked about examples recently. We can reference a few different things. We want to explain with this one that whatever you do, this is only about visual stuff.
matatk: Let's see if we can progress this a bit and come back to it in the next week or so.
matatk: Anyone that wants to join in may do so, just request edit privs and off you go.
CSS Selectors Level 4
<matatk> tracking: w3c/
hmat: Thanks Paul!
Issue Tracking
CSS Fonts - <meta text-scale> limits
<matatk> tracking: w3c/
matatk: The Web doesn't adapt to system font sizes. The reason is that it would break most websites. BUt if they opt into this caling behaviour, they are saying "I can cope with a a vabiable font size". What is said here it that it may only work up to a certain font size.
mat: The proposal is to say what the maximum font size copable in the value of this meta tag.
Fredrik: That should be matatk...
PaulG: There may be an incentive to race to the bottom here, only between those two sizes for instance.
PaulG: There is an issue where I have heading and body text. I need to read the body copy so I have my font size increasd on my device. The heading text may be already big enough for me to read. The reading text may then wrap and push everything down and makes everything a mess. Thi is where sophisticated sites have done better in taking al l
things into account. It isn't easy to put togetehr but works great Ths is going to be the poor man's scaled typography setting.
jan: I think we need to ask what support means. This is an interesting use case you laid out here, Paul.
<matatk> For reference: TAG design review of this proposal: https://
janina: The kinds of tconcerns Paul was raising: we need to be supporting tewaking as much as teakably tweakable.
matatk: I you've desined yoru site to be repsonsive to ext scaling, it is more likekly to work. I wouldn't be encouraging people to limit it to a specific thing.
PaulG: The typography is going to difer betwteen the platofrm and the OS. How iOS handles scaling between heading and body copy may diffeer from Android etc.
matatk: Ensure that the site can handle it, provide an excape hatch for the user if it can't and look at the possibility for the browser being able to check for if it's broken or not.
matatk: I need to draft something here...
[PROPOSAL] Exploring Generative UI
<matatk> tracking: w3c/
matatk: Roy is also the staff contact for the Web and AI IG.
matatk: He has made a lot of proposals. Fro instance exploring gneerative UI,
matatk: There is a healthy lot of discussion there.
matatk: I for one will be adding my two cents. You should, too.
[PROPOSAL] Tracking and Exploring AI and Accessibility
<matatk> tracking: w3c/
matatk: There's a similar one about AI and a11y. Whcih got a comment from Leonie today.
<PaulG> webmachinelearning/
<matatk> w3c/
<PaulG> webmachinelearning/
PaulG: Look for the WebMCP, #65.
PaulG: I was encouraged last week that multiple projects in browser automationand testing have con erged on the a11y tree bieng a more efficient mechanism to get thigns into AI content.
janina: I was in an informationaarchitecture meeting last weekend. There was talk about that you have to write your web content for humans and machines. There are ways you write content to emphasize your SEO to be included.
matatk: Yes, it's great if the semantics helping the people actually help the machines. But people have used a11y information to sutuff things in that should be caught by the machines.
matatk: Thanks everyone for coming!