– DRAFT –
PEWG

Multi-pen support and persistent pointerId #353 w3c/pointerevents#353

ACTION: Rob to speak to Microsoft about using a more generalised structure for the deviceId info

<flackr> https://groups.google.com/a/chromium.org/g/blink-dev/c/3eU-AHH8x4k/m/oCO7WKt1AAAJ?utm_medium=email&utm_source=footer

Rob: asked questions in the intent thread to bring back to discussion

Rob: i think there's answers here, but we need to figure out WHEN the field is set. value is pseudorandom, equivalent to pointerid, but should it be set for cases where multiple pointers are set on same device (e.g. hand with different touches). should probably only have it when the system itself can differentiate when it's an input that only has one pointer

Olli: what is a device? it's kind of vague. maybe all touches are the same, as they have different pointerid

Rob: yeah, you can use a combination of pointerid and deviceid, to differentiate

Rob: i think that's how it's coded in chromium ...

Rob: but this needs to be called out / spec'd

Rob: explicitly asked in the latest for them to comment back on thread

Rob: don't have concerns about deviceProperties as the name, seems reasonable. can always change later...

Rob: i think we have enough to move forward on this

Olli: will then need to see how this affects click event

Olli: what happens if triggered by keyboard

Rob: probably same logic as we do for keyboard...

Olli: pointerid is -1 for keyboard

Update https://github.com/web-platform-tests/wpt/blob/master/pointerevents/META.yml #498 w3c/pointerevents#498

<smaug> web-platform-tests/interop#628

Olli: posted this, will be updated automatically

interop pointer events issue

Olli: test change proposal for pointer events web-platform-tests/interop#628

Mustaq: proposed changes...

Olli: items that don't currently have include/exclude ... should I go in to change them in spreadsheet? just comment there?

Mustaq: if you make changes in spreadsheet, then point it out in the comment on the issue there, so we can sync

Mustaq: one concern may be number of tests we want to include. too many?

Olli: think most of them are fine, some weren't in GH at all because of the issues with interop23

Meta-issue: update WPT to cover Pointer Events Level 3 #445 w3c/pointerevents#445

Mustaq: got rid of one issue, got question about 477

Mustaq: w3c/pointerevents#477 question if these relate to shadow DOM

Olli: definitely added tests for shadow DOM for various things

<mustaq> web-platform-tests/interop#380 (comment)

<mustaq> Do these two tests cover the missing shadow dom cases for #477?

<mustaq> https://www.irccloud.com/pastebin/taCxoszU/

pointerevent_pointer_boundary_events_after_removing_last_over_element.html

To be precise: the second promise_test

<mustaq> https://github.com/web-platform-tests/wpt/blob/master/pointerevents/pointerevent_pointer_boundary_events_after_removing_last_over_element.html

Rob: not sure if either of those deals with slotted element

Rob: because one of the cases in 477 we called out was if you have slotted...

Olli: maybe we need to point out in issue that this needs to also deal with the slotted case

Patrick: random question ... do we know if coalesced and predicted events have actually started to be used in the wild?

<flackr> getPredictedEvents use counter https://chromestatus.com/metrics/feature/timeline/popularity/2971

<mustaq> getCoalescedEvents use counter: https://chromestatus.com/metrics/feature/timeline/popularity/4598

<mustaq> Correction: the lat link is only for insecure context usage