Meeting minutes
Agenda Review & Announcements
janina, does anyone have anything to add?
… beginning with Roy_Ruoxi for announcements, starting with Monday.
Roy_Ruoxi, I think I've already sent the invite. The background is W3C set up a Web & AI interest group, focused on AI related topics.
… for all W3C groups to discuss AI topics
… the first meeting topic will be AI and accessibility
… matatk will give a presentation on AI and accessibility.
<Roy_Ruoxi> https://
janina, as you bring that up, I'll clarify what an interest group is.
… (describes, here's a link: https://
… The Web AI group is currently gathering interests
… just noting this calling is on Monday, US Eastern 10am, and local time zones available in the calendar invite.
Roy_Ruoxi that sounds like a good explanation. Anyone interested may join the conversation.
<Roy_Ruoxi> https://
matatk Updated the APA participant's guide.
… all the processes updated for the new GitHub stuff, and also how to do a review, types of review, review etiquette, etc.
… expect at some point to update the wiki page, too, incorporating feedback. Seeking people to give feedback on the process, etc.
New Charters
matatk we do have a new charter. Devices and Sensors Working Group charter…
[I was disconnected from the server]
Devices and Sensors Working Group 2026 Charter
<Roy_Ruoxi> - issue: w3c/
<Roy_Ruoxi> - charter: https://
matatk checking if there are any new deliverables… they've renamed "device orientation specification" to "device orientation and motion."
… there's a tentative deliverable of "web serial API."
… all seem like a natural evolution of what the group has done before, so I don't have any concerns for this.
janina Any comments, questions, objections?
… seeing none…
Spec review requests
matatk using sub topics prepared
Accessibility Implementation Guidelines (DVB BlueBook A185)
matatk DVB implementation guidelines
… we discussed last week, unsure if there are any developments since.
… awaiting some combination of me and janina to put in some issues on it.
… looking at the relationship to the MAUR, and I believe people volunteered for this. I need to ping you on it.
<matatk> w3c/
janina it's getting urgent at this point, so we don't want to delay this much longer.
CSS Anchor Positioning Level 1
matatk if there are things to be raised, please do quickly. There's a link at the top of the tracking issue to the instructions in the wiki.
<matatk> w3c/
matatk so, anchor positioning… a few people volunteered. we discussed last week. Marcelo had proposed some accessibility relevant scenarios.
… the plan is for me to make a document, collaborate on it a bit, then bring back to the group.
… maybe by next week's call we'll have something. The week after I'm not here, so that would be a good idea. I'll do a first draft to give (you?) something to respond to.
… have not yet made a tracking issue for this.
… if people are interested in reviewing this, I'll make a tracking issue and assign you to that.
… they've filled in the FAST checklist, looking for updates, particularly the "forced colors mode."
… as that feature is highly tied to contrast, particularly on Windows.
… there is a change log, changes around forced colors mode. any CSS enthusiasts up for this?
css-color-adjust-1
<matatk> w3c/
janina anything "forced" makes me nervous.
matatk this is about the accessibility of the feature, high contrast interaction with forced colors. there is guidance.
PaulG this is about a handshaking between authors request to preserve certain color combos for, say, their logo, branding, etc.
janina that's a good answer that gives me comfort.
… I'm obviously not a user of this, but the author proposes is my guiding principle of any of these display characteristics. but the user still disposes... I can't use that, I need it this way.
PaulG such as hazard placards that belong on caustic materials, you can't just force colors, it needs to match the circumstances.
matatk question for PaulG: is this related to or overlapping with "only light" or "only dark" in the color scheme property?
PaulG there's some things in here about the detection of lightness by the user agent. And to determine which bucket a page or element falls in, and which color scheme.
the. It says here that for pages that do not explicitly support dark color scheme, and have not explicitly forbidden an auto adjustment by specifying color scheme. Only light.
Then this mode, triggers, auto adjustments. So there's like.
There's some things that authors are gonna have to do.
In their css to protect certain things. And then there's also things they're gonna have to do in the meta of the page to say, you know this page needs to.
… authors will just have to be aware of this.
… now that user agents have this capability, people will need to review published material to make sure it meets these requirements.
matatk has this conversation interested anyone in looking at it?
Steve Faulkner expressed interest.
matatk that sorts that,
… this is to do with data shapes, which is an RDF thing.
… they've recently done a new version of RDF.
… which is this SHACL thing, some sort of querying or processing thing on top of RDF. They've modularized the spec.
… they list all the different specs that relate to SHACL in different areas.
… one area is about user interface generation, but that's not this one.
… I am interested as that was my area of research.
… but this isn't that one. If anyone wants to volunteer to check on this…
SHACL 1.2 Core
<matatk> w3c/
Neha volunteers.
matatk Thank you, I'll make a meta thread.
PaulG I'm noticing this loaded for me in dark mode.
@q
… it was set to auto, but it did dark mode. And the first instance of JSON has such low contrast like 1:2.
matatk this is interesting, because we did have this issue where the dark mode colors for Respect by default have some contrast issues.
… I need dark mode as light mode hurts me.
… so, if this is vexing people, if you want to help.
<PaulG> got to https://
@PaulG and I both have the astigmatism issues with this ReSpec's dark mode issues.
matatk there's two separate issues we're talking about: the default color palette issue, and the other the one we're mainly talking about is the dark/light mode issue where it is still sticking. I don't know if there's an existing issue for that. If there is, let's pile on it. If there isn't, let's make one.
… does anyone have cycles to look for a Respec issue? If so, email the APA public list, and we'll take it from there.
janina We can initiate a concern.
Respec issues: https://
I opened the issue: speced/
We've had people need dark mode because light mode hurts, others find dark mode's current palette causes issues for those with astigmatism, and then the lack of stability of choosing a preference and getting the selected creates confusing for the user that becomes distracting for the work at hand.
Issue tracking
Coordinates of keyboard-driven click events
<matatk> w3c/
matatk It's asking for... it's talking about. the coordinates of keyboard-driven click events.
And it's an issue on HTML. which was filed recently, I think, because this is quite new, so it was open 5 days ago by Anna Van Kestern from Apple.
Um, and... Uh, it is... Um... talking about what those coordinates should be.
for events... click events that come from the keyboard.
And, um, there's a couple... of, um... observation, so the defaults would be, uh, if a mouse is never moved by the user, assuming the mouse is never moved.
matatk the coordinates for the click event would be 0, because, um, that's.
kind of inherited from where the mouse position is.
But Anna is saying that this is, uh, found to be bad for accessibility, because menus can end up being positioned wrongly as a result, for example.
Patrick has, um... commented. Uh, so has several other people.
There's quite a bit of a thread. They're developing.
Um, so... Uh, we... it may well be the case that we can let the discussion go on, and it's all in good hands and everything.
Which it seems to be, but it'd be nice to have a read.
on where it looks like it's going, and whether anyone thinks APA should.
say anything about this in particular. If anybody has opinions about this particular issue.
Uh, so, does anyone want to volunteer to at least.
Um... go and look and see... what the.. the problem is, and... and... how it looks, this current state of it, the issue.
PaulG How would you trigger a mouse event. With the enter in space bar that that shouldn't trigger the mouse event or pointer event. It should be triggering.
Click and keyboard, and and key down and key up and press.
matatk Click has mouse coordinates associated with it.
PaulG and matatk are both checking MDN.
matatk It's a pointer event that gets passed in.
<matatk> We looked at MDN for the click event - https://
PaulG has put this one on his list.
Everyone sends Paul and his partner best wishes.