Meeting minutes
<janina> /join #rqtf
Agenda Review & Announcements
janina: A bit of buzz regarding AI agent that is frontending various kinds of purchases.
<janina> https://
janina: Tries to stick to the essentials and avoid the ancillary distracting and complicating things people often complain about
janina: It's a personal agent, but wouldn't call it an overlay
Fredrik: Could we discuss this at a point?
janina: RQTF will discuss around accessibility of AI.
matatk: If it's going to, we should post to APA in case anyone can attend that RQTF meeting and contribute
Fazio: It sounds a lot like an overlay, but it gives that impression that it is that kind of technology. Worth exploring.
WebExtensions WG (to be?)
<matatk> https://
matatk: Is a community group called Web Extensions which is about browser extensions
matatk: Mostly followed by the schema put in place by Chrome
matatk: Parts of extensions get access to privileged APIs.
matatk: Other parts exist in the browser process and don't interfere with the page's DOM but can access APIs (open sidebars, change keyboard shortcuts, open settings, etc.)
matatk: Community group exists to keep as much harmony as possible within the model of extensions. Several issues that affect everyone who writes and consumes extensions. This group is doing a lot of work to make things viable and uniform and solve these issues.
<Fredrik> Any chance that one could tweak Cisco by calling it the Webex WG?
matatk: We were approached by Simeon (Mozilla dev). Asked APA about becoming a WG. Not sure we'll need an official liaison between the two groups.
matatk: Wanted to highlight this and encourage questions and concerns
janina: A gracious response, Matthew. We care a lot about extensions but it's worth thinking about a liaison.
APA re-charter
<matatk> w3c/
matatk: Some questions about process documents, are we going to work on them. Discussed with Roy last week. Roy has responded and making appropriate changes.
Roy_Ruoxi: W3C review has finished. I started request for review. Sorting out logistics for this process. Things will start from tomorrow for the AC review.
Roy_Ruoxi: Will require at least 28 days
TPAC 2025
matatk: Confirm that we put in requests for lots of meetings with specific groups. All have been received. Scheduling team is hard at work to organize.
matatk: Breakouts yet to be scheduled.
mike_beganyi: funding opportunities?
matatk: Fee waivers for invited experts in the past
matatk: Been long-term request to provide funding for IEs that people could apply for. Being looked into, but no status update at the moment.
janina: Recent years there has been some funding. Presume it'll be made available again. This year is more challenging due to location.
Fredrik: Last year there was something e-mailed out in that respect
Fazio: Closer to TPAC will be sent out via e-mail
Horizontal Review process update
matatk: Updates for carrying out reviews.
matatk: Available information on the Wiki. Will be sent to the list at the end of the meeting.
matatk: [summarizes wiki page and changes]
<matatk> https://
New on TR
Roy_Ruoxi: Nothing this week.
Charter review requests
Roy_Ruoxi: Nothing this week.
Spec review requests
matatk: A couple of issues here. Will signpost them.
matatk: Point Jennifer made regarding changes to CSS interactivity property and it has a value of inert
matatk: All part of wider project in CSS seeking to make it easier for web devs to make carousels.
matatk: Two design reviews from the TAG. Agree with Jennifer's concern. Some of the inert stuff we're more concerned about has been paused.
matatk: Second item is web neural networks API. Filed comments we had last week into one big issue (which we won't do anymore). Were two extra things about this. We could make these comments as separate issues.
matatk: Could easily propose a comment in APA repository to file as an issue on that spec. Paul's question of adding things to the accessibility considerations section of specs.
CSS Update (Paul)
Scroll markers
<PaulG> w3c/
PaulG: Scroll markers want to add roles to something. Consult the examples on MDN it's all carousels.
PaulG: What they're adding is tab or link as an implicit role in the scroll marker.
PaulG: This is strange. Developers without context won't get the deeper concerns of this method.
matatk: Is there a reliable way for the browser to know which elements are being presented with various roles via CSS?
PaulG: CSS is providing the "target" and that's what's making the difference for interpretation.
PaulG: Should look at prototypes and then raise concerns
PaulG: It's not always there. They have to pick it and discern the best option.
matatk: What about the behaviour? Some of CSS stuff imparts behaviour, like keyboard handling.
PaulG: Not sure this is clear in the proposal. As they're prototyping maybe this will come up
matatk: Odd that you would get behaviour out of CSS as opposed to custom elements. CSS does behaviour now? Huh?
PaulG: Seems they're trying to stretch the function of CSS
Animation Trigger
<PaulG> w3c/
PaulG: Has to do with parallax scrolling techniques and make these more native without relying on JS. Animation triggers to have play and pause with scrolling intent. Trying to figure out API
PaulG: Making this native doesn't seem like the best plan as some users will have negative reactions to this for various reasons, disability included.
PaulG: This can be done today, but not with CSS alone.
matatk: This also seems concerning.
matatk: Will see if CSS and a11y folks on TAG are aware of it and how to respond.
PaulG: I'm more concerned about pointer than I am about scroll.
PaulG: A couple of methods in scroll to mitigate this, but pointer...so many people have fine motor issues and this could cause significant challenge
matatk: Does this need to be separate or can we keep it in Prefers Reduced Motion
PaulG: Could be in Preferred but it could be its own thing.
Comment review requests
Canvas Text Metrics for Editing, Art and Design
<matatk> source: w3ctag/design-reviews#1095
<matatk> tracking: w3c/
matatk: Came to the TAG as a design review. Early proposal.
matatk: Allows layout of text on a canvas that follows a particular path.
matatk: Concern is about the accessible fallback. How do you get access to that text if it's important? Currently, the only way is to provide canvas fallback via the canvas element. Could be part of Accessibility Object Model in future
matatk: Has been some discussion on the topic. Worth investigating and leave a comment.
looks cool