Pointer Events WG Voice Conference -- 05 Feb 2013

<scribe> Scribe: Art

<scribe> ScribeNick: ArtB

Date: 5 February 2013

<mbrubeck> Good morning, Art.

Agenda

AB: I posted a draft agenda a couple of days ago http://lists.w3.org/Archives/Public/public-pointer-events/2013JanMar/0061.html. Since then there has been a relatively significant amount of activity on various agenda topics.
... we could now try to sort out what needs to be dropped, changed, added; or, it could be more time efficient to just stick with the draft agenda and adjust it accordingly as we proceed.
... we can also add a discussion about LC plans to the AoB section.
... is that OK? Any other proposed additions

JR: want to walk through some of my e-mails

AB: sure, that's fine

Making click/contextmenu use PointerEvent interface

AB: this topic was raised by Rick a few weeks ago http://lists.w3.org/Archives/Public/public-pointer-events/2013JanMar/0024.html and we talked about it last week http://www.w3.org/2013/01/29-pointerevents-minutes.html#item02.
... it appears we still need feedback from Rick so I propose we postpone this topic and I can assign an Action to Rick to reply to this thread. OK?

JR: I think that's fine

… I raised this issue and if it takes some more time that's fine

… I dont think it will require changes to PE spec

… but perhaps some spec in HTML and/or WebApps

<scribe> ACTION: Rick reply to the "Making click/contextmenu use PointerEvent interface" thread [recorded in http://www.w3.org/2013/02/05-pointerevents-minutes.html#action01]

<trackbot> Created ACTION-18 - Reply to the "Making click/contextmenu use PointerEvent interface" thread [on Rick Byers - due 2013-02-12].

Mapping for devices that don't support hover and CSS :hover; bug 20222

AB: the hover issue is documented in bug 20222 https://www.w3.org/Bugs/Public/show_bug.cgi?id=20222 and Rick started a related thread http://lists.w3.org/Archives/Public/public-pointer-events/2013JanMar/0041.html.
... an issue is whether or not the current text regarding hover is overly restrictive (given hover's usage `in the wild`). And there is also an issue about making sure v1 doesn't preclude us from doing

… something about this in v2

AB: we postponed this topic last week because Rick was not present. He followed up yesterday http://lists.w3.org/Archives/Public/public-pointer-events/2013JanMar/0067.html and then Jacob replied.
... it appears the way forward is: a) to leave text as is for v1; and b) to add hover to the list of potential features for v2.

JR: I stated the right approach on the list

… think some experimentation is still needed here

… want a general solution

… Rick and I agree this is a tough problem to solve generally

… and that doing something in v2 may be the right thing to do

<mbrubeck> I also support punting any changes to v2.

MB: I support leaving current text as is and potentially doing something in v2

SG: agree

DS: fine with that but it should be called out because it is an Accessiblity issue

JR: that's fair

… I can add a related note

… identify the problem but note we don't have a solution

… Hover has always been an a11y issue

… It is worth noting PE doesn't fix this issue nor make it worse

AB: so Jacob will add some non-normative text

… and also close 20222

… is that correct?

JR: yes

<scribe> ACTION: Jacob close bug 20222 per the 5-Feb-2013 resolution [recorded in http://www.w3.org/2013/02/05-pointerevents-minutes.html#action02]

<trackbot> Created ACTION-19 - Close bug 20222 per the 5-Feb-2013 resolution [on Jacob Rossi - due 2013-02-12].

<scribe> ACTION: Jacob add some non-normative text for 20222 [recorded in http://www.w3.org/2013/02/05-pointerevents-minutes.html#action03]

<trackbot> Created ACTION-20 - Add some non-normative text for 20222 [on Jacob Rossi - due 2013-02-12].

Transformed Pointer Coordinates?

AB: Doug Schepers started the conversation via http://lists.w3.org/Archives/Public/public-pointer-events/2013JanMar/0046.html. Based on that discussion, Jacob opened a bug for the CSSOM View https://www.w3.org/Bugs/Public/show_bug.cgi?id=20832 spec.
... it appears there is nothing we need to do for the PE spec. Is that correct?

DS: yes, that's correct

… the result is that people have some ideas

… they agree it is not a PE issue but an issue for CSSOM spec

RESOLUTION: the "Transformed Pointer Coordinates" issue will be addressed in other specs (not the Pointer Events spec)

PE spec comments by Cathy

AB: Cathy asked some question about the PE spec http://lists.w3.org/Archives/Public/public-pointer-events/2013JanMar/0057.html
... Jacob replied yesterday http://lists.w3.org/Archives/Public/public-pointer-events/2013JanMar/0072.html. He suggests two comments are mostly editorial and he added two new bugs:
... 1. "PREVENT MOUSE EVENT flag should be per pointer type" https://www.w3.org/Bugs/Public/show_bug.cgi?id=20872
... 2. "Pointercancel should also implicitly release capture" https://www.w3.org/Bugs/Public/show_bug.cgi?id=20873

JR: this was good feedback

… I can walk through them

… #1 how many pointers can be the primary pointer at once

… there can in theory be many devices

… but in reality, don't see many devices used at once

… End up having mouse events fighting each other

… We considered device arbitration

… but we didn't go that route

DS: if a game has a tablet and 2 people and each person has an input device - does this work?

JR: absolutely

… primary gives UA a rule for multitouch to determine which device wins

DS: do you have some code to demonstrate this? If yes, that would be great to show during W3Conf

JR: ok; I can look into that
... if anyone has feedback on other scenarios like that, please send email
... ok, Cathy's comment #2

… the spec will need to be changed

… it's a simple change

… filed bug 20872

… https://www.w3.org/Bugs/Public/show_bug.cgi?id=20872

… I proposed changes in http://lists.w3.org/Archives/Public/public-pointer-events/2013JanMar/0072.html

<Cathy> Looks fine to me.

AB: any comments re Jacob's proposal for #2?

JR: re Cathy's comment #3 ...

… agree the text is a bit convoluted

… I can clarify this text

… I didn't open up a bug because the fix is relatively easy

JR: re Cathy's comment #4

… there is a scenario missing from the spec

… for pointer cancel

… I opened bug 20873 for that https://www.w3.org/Bugs/Public/show_bug.cgi?id=208723

… will make a simple text change to fix this

AB: sounds good to me

<asir> Correct link is https://www.w3.org/Bugs/Public/show_bug.cgi?id=20873

CC: yes, that's good

Pointer events – active buttons state & pen devices

AB: Tim Bannister asked about how PEs work with pen devices and buttons http://lists.w3.org/Archives/Public/public-pointer-events/2013JanMar/0052.html
... Jacob replied yesterday http://lists.w3.org/Archives/Public/public-pointer-events/2013JanMar/0069.html.
... it appears this was mostly a request for clarification and that no spec change is needed. Is that correct?

JR: yes

… I haven't seen the behaviour Tim described

… the button is just a modifier

AB: any comments?

… ok, then we consider that resolved

Pointer Events Open Bugs

AB: the only open bug http://tinyurl.com/Bugs-PointerEvents we haven't talked about is 20109 that Jacob reopened yesterday https://www.w3.org/Bugs/Public/show_bug.cgi?id=20109
... Jacob's comment http://lists.w3.org/Archives/Public/public-pointer-events/2013JanMar/0073.html

JR: this is the mouse emulating pressure bug

… we realize that in most pen apps, for max pressure for the pen correlates to greather than normal stroke width

… half-pressure would be "normal" pressure

… if use something like Msft Paint, a line is .5 pressure

… We changed the spec about a month ago

… changed from .5 to 1

… Need to think about other properties e.g. tiltX, tiltY, width, height

… Can leave it up to impls to decide

… If we emulate one prop, we should emulate the other properties too

… We are proposing : mouse should be .5 rather 1 for pressue

… spec now use MAY for width and height and we think it should be SHOULD

MB: if max should be twice default pressure but in Msft doc it's 1.5 times

JR: the documentation is correct, I made an error in my e-mail

MB: ok, thanks

… I agree with the proposal to change this to SHOULD

… we need consistency with the defaults

… otherwise, get interop problems

… so having consistent defaults is a good idea

AB: any other comments

<mbrubeck> Rationale: If developers test only in browsers with a default value, they might not realize their program has a divide-by-zero bug or other bugs, in browsers without a nonzero default.

AB: so do we ask Jacob to change this now?

JR: I can make the change and then ask Rick and Olli if they have any concerns

<scribe> ACTION: Jacob apply your proposed change to bug 20109 and ask the group for comments [recorded in http://www.w3.org/2013/02/05-pointerevents-minutes.html#action04]

<trackbot> Created ACTION-21 - Apply your proposed change to bug 20109 and ask the group for comments [on Jacob Rossi - due 2013-02-12].

<asir> https://www.w3.org/Bugs/Public/show_bug.cgi?id=20217

AB: anything else on the spec before AoB

JR: re touch-action

<asir> You can find Rick's response http://lists.w3.org/Archives/Public/public-pointer-events/2013JanMar/0067.html

… Rick replied and was OK with the proposal

Any other Business

AB: re LC, it appears we have 3 open bugs (20109, 20872 and 20873). If we still want to get a LC published on Feb 19, we should be in a position to discuss this transition during our Feb 12 call. That means we need proposed fixes for these bugs `real soon now`.
... I think we are in pretty good shape

AV: we are still waiting for Rick's input on the context menu issue

AB: oh yes, good point

DS: can someone make sure Rick knows that?

AB: I already have an action to contact Rick

DS: would be good to get the bugs fixed as soon as we can

JR: yes, I'm on it

AB: if you have any comments regarding the IndieUI: Events 1.0 FPWD http://lists.w3.org/Archives/Public/public-pointer-events/2013JanMar/0059.html, please send them to public-indie-ui-comments @ w3.org.
... and news regarding implementations?

AV: call next week?

AB: yes, to record consensus to publish LC

DS: re W3Conf ...

… if people have implementation info we can share, that would be good

… would be good if JQuery would create a blog about their plans related to Pointer Events

… then Jacob could mention that during his presentation

… it would send a good message

SG: yes, I'll talk to others in jQuery

DS: are there any especially cool demos that Jacob could potentially use during his preso?

JR: yes, proposals welcome [and slides ;-)]

DS: think a two-player game would be nice

<mbrubeck> Air hockey, e.g.

JR: we worked with Atari on some games like pong and those games use pointer events

AV: what about a call on Feb 19?

AB: yes, that's fine with me
... so we will have a call next week if there is something to discuss other than a CfC to publish LC

… if LC is the only topic, I will use email and a short review period - like 1 or 2 days

AB: meeting adjourned

- DRAFT -

Pointer Events WG Voice Conference

05 Feb 2013

Attendees

Contents