Minutes from PEWG meeting 22 October 2025

Dear all,

the minutes from today's meeting are at 
https://www.w3.org/2025/10/22-pointerevents-minutes.html and copied below.


PEWG
22 October 2025

Agenda: 
https://www.w3.org/events/meetings/66591f6b-6694-4f90-b23d-bf8f1b9dda8a/20251022T110000/
IRC log: https://www.w3.org/2025/10/22-pointerevents-irc

Attendees
flackr, mustaq, Patrick_H_Lauke, plh, smaug

Chair: Patrick H. Lauke
Scribe: Patrick H. Lauke


* Web Platform Tests https://wpt.live/pointerevents/ / 
https://wpt.fyi/results/pointerevents?label=experimental&label=master&aligned

* what to do with UI events



# Web Platform Tests https://wpt.live/pointerevents/ / 
https://wpt.fyi/results/pointerevents?label=experimental&label=master&aligned

Patrick: on this topic, i spoke to PLH checking what we need to do to go 
to REC with Level 3, and he reminded me kindly that ideally we need two 
passing implementations for each feature. question: if we have some WPTs 
that fail because the test harness doesn't support things, is that a 
problem?

<mustaq> I found this to be a test problem (vs impl): 
pointerevent_after_target_removed_from_slot.html?mouse

PLH: ideally have passing result, but it can also be part of the "story" 
- if tests fail because of limitations of the test platform, and we feel 
that things are as good as they can be in terms of "it's 
implemented/correct, just that the test is flawed/we can't test for it", 
then we need to document THAT

PLH: so you believe all features are implemented, regardless of tests?

Olli: i think the new touch-action values are chromium only

PLH: my recommendation then would be to drop those if we don't have two 
implementations. dropping would allow us to ship level 3. they're not 
marked as at risk, so may need to issue another CR

<mustaq> New touch-action values are unidirectional ones (left, right vs x)

PLH: you would need to drop those from the PE3 draft (in the branch)

ACTION: Patrick to remove the chromium-only touch-actions from the level 
3 branch for a new snapshot

<mustaq> 
https://wpt.fyi/results/pointerevents/pointerevent_touch-action-pan-x-css_touch.html?label=experimental&label=master&aligned

<mustaq> 
https://wpt.fyi/results/pointerevents/pointerevent_touch-action-pan-left-css_touch.html?label=master&label=experimental&aligned

Mustaq: firefox fails pan-x the same way as pan-left, so this might be a 
harness problem? (as pan-x at least should pass)

Patrick: if there are ones that fail because of harness, like the touch 
ones, can we do a manual test instead?

Olli: there are also failures from pointerlock ... (which might be 
weirdly intertwined with pointer events?) they seem to all be failing at 
the moment

Mustaq: i think those were passing before in chromium...but they're 
covered by pointerlock

Olli: yeah just wasn't sure if there was overlap with pointerevents

Rob: it's not even part of the pointerevents spec

Mustaq: some of them are indirectly dependent on pointerevents ...

Patrick: the tests might be failing because of something that changed in 
pointerevents, but not directly our concern. needs to be looked at by 
pointerlock folks

Patrick: suggest at TPAC we try to make a big list of tests that are 
problematic/may need to be changed to manual tests



# what to do with UI events

Patrick: [explains the situation: UI Events is stale, has vague 
definitions, but we now come across things that we'd want to specify 
more precisely ... but how?]

PLH: it does indeed look like nobody is actively working on UI events. 
but in first instance, to avoid monkeypatching, suggest sending a PR to 
UI events ...

Rob: we tried at last TPAC to get clarity, but no joy since

<mustaq> See Day3 in 
https://docs.google.com/document/d/1EtEFfQ3sJ6lECF3ojmbLKIH7iKN4RdGrapVosQB-wqI/edit?usp=sharing

Olli: or we just remove things from UI events and spec them ourselves...

PLH: or that yes ... send a PR that removes things from UI events and 
migrates definitions to PE

Rob: it's not that there was opposition, just ... nobody to do it

PLH: my suggestion: send PR to UI events pointing to PE instead of vague 
definiton. would guess it impacts all of section 5.3 in UI events

Olli: we should probably also move mouse events. so UI events becomes 
keyboard events and composition events

PLH: also wheel events

Rob: wheel events is probably another pointer event that we want to 
capture perhaps

<plh> https://www.w3.org/TR/uievents/#external-pointerevent-algorithms

Mustaq: that note was probably added just before last TPAC

PLH: need to see if any more things need to be moved, like 
mousedown/mouseup/etc. we ready to do the work on it yet?

Mustaq: not yet

PLH: if we are going to DO the work, then we need to extend the charter 
to mention mouse, wheel, etc

Olli: UI event spec is also in a limbo state where it started having 
algorithms/pseudo-algos, but not quite yet

Patrick: of course, this would be for Level 4, not 3...

Olli: no spec has algorithms for certain things YET, so we need to come 
up with them first before sending a PR

ACTION: Patrick to reword/expand charter to include mouse events/pointer 
events algo (potentially wheel events?)

Patrick: pledging that i will have the PR for the charter by next 
monday. We'll reconvene in 2 weeks' time before TPAC, and then at TPAC 
we have the meeting on Friday. thank you all


Summary of action items

* Patrick to remove the chromium-only touch-actions from the level 3 
branch for a new snapshot

* Patrick to reword/expand charter to include mouse events/pointer 
events algo (potentially wheel events?)

-- 
Patrick H. Lauke

* https://www.splintered.co.uk/
* https://github.com/patrickhlauke
* https://flickr.com/photos/redux/
* https://mastodon.social/@patrick_h_lauke

Received on Thursday, 23 October 2025 19:20:09 UTC