Minutes from Pointer Events WG call 21 March 2018

Minutes from today's call:

Pointer Events WG
21 Mar 2018
Attendees Present: NavidZ, Olli_Pettay, Patrick_H_Lauke, Mustaq
Chair: Patrick H. Lauke
Scribe: Patrick H. Lauke

<NavidZ_> 
https://github.com/w3c/pointerevents/issues?page=1&q=is%3Aissue+is%3Aopen+-label%3Afuture-v3&utf8=%E2%9C%93

<NavidZ_> 
https://rawgit.com/smaug----/test-results/d2b1bcc451c5b85d2b1d0cb721a2f732b3c01120/pointerevents/less-than-2.html

Above link is what Google care most about in Olli's test/PR

Olli: apparently some tests were missed. Framework for (manual) tests is 
broken/times out

<NavidZ_> https://github.com/w3c/pointerevents/issues/229

Patrick: we'll go through issues

#229 future-v3

<NavidZ_> https://github.com/w3c/pointerevents/issues/226

#226 assigned to Navid. we discussed last week, let user agents handle it

<NavidZ_> https://github.com/w3c/pointerevents/issues/225

#225 same - discussed last week, added test for that yesterday

<NavidZ_> https://github.com/w3c/pointerevents/issues/221

#221 discussed last meeting, asssigned to Patrick

<NavidZ_> https://github.com/w3c/pointerevents/issues/220

#220 we're just going to follow terminology for pointer lock

<NavidZ_> https://github.com/w3c/pointerevents/issues/219

#219 assigned to Olli

<NavidZ_> https://github.com/w3c/pointerevents/issues/202

#202 new one, not discussed yet. assigned to Patrick atm. this comes 
from direct manipulation inputs that scroll.

do we want different pen actions, or do touch actions also apply to pen 
actions?

chrome currently: if the direct manipulation device is capable of 
scrolling (see input capabilities), it listens to touch-action

Navid: Olli will firefox do something here do you know? Olli: no input 
currently on this one

Mustaq: we may defer to v3, as we haven't got consensus yet

Patrick/Olli agree

<NavidZ_> https://github.com/w3c/pointerevents/issues/183

#183 assigned to patrick

<NavidZ_> https://github.com/w3c/pointerevents/issues/177

#177 Olli: there's a test for this, but Firefox is failing. we have some 
inconsistency which events are logged and which are not

(w3c tests)

<smaug> 
https://searchfox.org/mozilla-central/source/dom/html/nsGenericHTMLElement.cpp#2222

Navid: should we define this in PE?

Patrick: would prefer having it in upstream (e.g. UI Events?)

Olli: Chrome was going to change this earlier this year - mouse events

Navid: should we fire events then on disabled?

Olli: certainly not click, chrome currently fires pointerup and pointerdown

Navid: in latest stable, chrome sends all pointer events, but no mouse 
events

<NavidZ_> https://www.chromestatus.com/features/5685077795143680

Navid i'll add this to myself, look for the test, expect 
pointerup/down/move on disabled events, but no click

Patrick: should this be mentioned in the PE spec, or is it already 
implied/mentioned in other upstream specs?

<NavidZ_> 
https://github.com/w3c/web-platform-tests/blob/master/pointerevents/pointerevent_disabled_form_control-manual.html

Olli: it should be said explicitly *somewhere*

maybe just a note somewhere about disabled elements

navid: i'll add a note as well to spec

<NavidZ_> https://github.com/w3c/pointerevents/issues/173

#173 question: when we release pointer capture, should boundary events 
be fired/update hover etc

when capture is release, in chrome at least, it waits until NEXT event 
to then fire all the events to get out of the previous (captured) 
element and into the new element the pointer is now over

chrome does currently for concrete events. olli do you know how firefox 
behaves?

olli: don't recall right now. thinking what right way to do it would be

<smaug> I lost audio

navid/mustaq discuss interaction between scrolling by keyboard and 
boundary events (as page and mouse move against each other)

(smaug shout if you get back in)

navid/mustaq discuss pointercapture - that after getpointercapture 
chrome also waits until next concrete event

(setpointercapture)

boundary events are equivalent to hovering - after pointer leaves, page 
cannot find out which element is currently hovered until next concrete 
event?

Olli: yeah agree that makes no sense. hover is always updated

Navid: so you agree we SHOULD fire boundary events right away after 
pointer capture is released?

Olli: yes, that would be what developers expect I think

Navid: chrome currentyl waits for next concrete event, but sending right 
away makes more sense

Olli: i think Firefox sends mouseover when page scrolls using keyboard, 
but not mousemove
... would be good to test current behavior

Navid: Olli can you take care of testing?

Olli agrees

Patrick: and add a note in spec clarifying this

<NavidZ_> https://github.com/w3c/pointerevents/issues/172

#172 dupe of https://github.com/w3c/pointerevents/issues/219?

https://github.com/w3c/pointerevents/issues/171

#171 navid: i believe dblclick itself is kind of legacy, as click 
received the detail/click count

navid: seems outside of scope of pointer events. and that the issue 
itself is a mess. but we never talked about it in PE. it should go 
somewhere else like UI Events

agree we should either close or move to another spec

<NavidZ_> https://github.com/w3c/pointerevents/issues/167

#167 Olli these are old IE things that were unspecified

Navid: oh i see, as these are updated in mouse events we'd inherit those
... should we call these out that we don't inherit them / set them to null

Mustaq: as we're forced to inherit them....

Navid: what's the use case? If authors were to migrate their old legacy 
code to PE, would they be able to do it in a more standard way?

If there's better way for authors (as they already need to update their 
code), then we should set them to null

Patrick: agree
... so are we going to add a note that we blank/null them?

Navid: yeah once we've looked at use cases

https://github.com/w3c/pointerevents/issues/165

#165 Patrick: not married to the idea

Navid: this is related to UA ignoring/cancelling multiple pointer events

maybe you can look at that note and see if it can be added there

Patrick: sure

<NavidZ_> 
https://rawgit.com/smaug----/test-results/d2b1bcc451c5b85d2b1d0cb721a2f732b3c01120/pointerevents/less-than-2.html

NAvid: would like to go over test results for rest of call
... new touch action behaviors in chrome, not in any other browsers

do we want to keep them?

Olli: should move them to v3

Edge doesn't have those either

Because I believe Edge's results only cover automated tests at this point

Navid: you do have the tangentialpressure attribute (but no value), right?

<smaug> https://bugzilla.mozilla.org/show_bug.cgi?id=1292064

Navid: you're passing a test we fail (movementxy). do you have 
fractional coordinates in Firefox?

we are failing it because pointerlock hasn't updated yet

Olli: yes we're failing that as well. need to revisit those results
... still waiting to get a pen device to run manual tests

Navid: we have 15 issues left, (and 25 assigned) so we can cover those 
next week

Summary of Action Items
Summary of Resolutions
[End of minutes]




-- 
Patrick H. Lauke

www.splintered.co.uk | https://github.com/patrickhlauke
http://flickr.com/photos/redux/ | http://redux.deviantart.com
twitter: @patrick_h_lauke | skype: patrick_h_lauke

Received on Wednesday, 21 March 2018 16:07:50 UTC