Draft mintes: 2014 March 25 call

The draft minutes from the March 25 voice conference are available at 
the following and copied below:


WG Members - if you have any comments, corrections, etc., please send 
them to the public-pointer-events mail list before April 1. In the 
absence of any changes, these minutes will be considered approved.

-Thanks, ArtB


       [1] http://www.w3.org/

                                - DRAFT -

                    Pointer Events WG Voice Conference

25 Mar 2014


       [2] http://lists.w3.org/Archives/Public/public-pointer-events/2014JanMar/0194.html

    See also: [3]IRC log

       [3] http://www.w3.org/2014/03/25-pointerevents-irc


           Arthur_Barstow, Jacob_Rossi, Asir_Vedamuthu,
           Scott_Gonzalez, Cathy_Chan, Rick_Byers, Olli_Pettay,
           Matt_Brubeck, Doug_Schepers

           Sangwhan_Moon, Patrick_Lauke




      * [4]Topics
          1. [5]Tweak Agenda
          2. [6]Bug 24923: What should happen to the mouse events
             if pointer event listener removes the target ...
          3. [7]Bug 24971: Should got/lostpointercapture be
             dispatched asynchronously or synchronously
          4. [8]Sub-pixel coordinate granularity
          5. [9]Touch-action to SVG elements
          6. [10]Exception usage
          7. [11]Feedback on pointer events
          8. [12]Testing: Jacob's latest commits for PR 324 needs
          9. [13]CR implementation updates
         10. [14]AoB
      * [15]Summary of Action Items

    <scribe> ScribeNick: ArtB

    <scribe> Scribe: Art

Tweak Agenda

    AB: any change requests to the draft agenda

      [16] http://lists.w3.org/Archives/Public/public-pointer-events/2014JanMar/0194.html?

Bug 24923: What should happen to the mouse events if pointer event
listener removes the target ...

    AB: we discussed but 24923
    <[17]https://www.w3.org/Bugs/Public/show_bug.cgi?id=24923> on
    March 11
    em07> and didn't reach consensus on what to do (or not do).
    There has been no followup in the bug or on the list since that

      [17] https://www.w3.org/Bugs/Public/show_bug.cgi?id=24923%3E
      [18] http://www.w3.org/2014/03/11-pointerevents-minutes.html#item07%3E

    OP: think we should follow IE

     and put that behavior in the spec

    RB: would be good to get details from Jacob

     and he did that

     now we need to spec it

    JR: agree we want that in the spec

     not sure how I would insert this into the spec

     could be better for me to create a changeset and for people
    to discuss that

    RB: sounds good

    <scribe> ACTION: jacob create a changeset for bug 24923 and
    sent it to the list for review [recorded in

    <trackbot> Created ACTION-98 - Create a changeset for bug 24923
    and sent it to the list for review [on Jacob Rossi - due

    AB: ok, thanks Jacob

Bug 24971: Should got/lostpointercapture be dispatched asynchronously
or synchronously

    AB: we discussed bug 24971
    <[20]https://www.w3.org/Bugs/Public/show_bug.cgi?id=24971> on
    March 11
    em08> and didn't reach consensus on what to do (or not do).
    There has been no followup in the bug or on the list since that

      [20] https://www.w3.org/Bugs/Public/show_bug.cgi?id=24971%3E
      [21] http://www.w3.org/2014/03/11-pointerevents-minutes.html#item08%3E

    JR: I can add some info to the bug

     Olli had a related Q re the order

     in our imp when calll setptrcapture, we set a pending capture

    [ Jacob gives details of IE impl  ]

     So this is similar to what Olli says in comment #2

    [ Jacob gives details about Olli's comment #4 ]

    OP: that gives a surprising result

    [ Olli gives a scenario that could result in surprising results

    JR: if get gotcapture, we guarantee you get a lostptrcapture

     I agree we need to update the spec

     Similar to one of Anne's comments

     Need to make the spec clearer but need to make sure we all
    agree on the behavior for these scenarios

    OP: wonder if this is just a bit too complicated and just fire
    synch events

     synch handling would make the algorithm simpler

    JR: we had probs with apps when we had synch firing of the

     we could say that's bad app behavior

     i.e. tight loops of setting capture

    RB: would like to understand what the apps were trying to do in
    those cases

    JR: that would require some investigation (been a few years)

    SG: are storage events synch (like localStorage)

    JR: if there are no other synch events, we probably don't want
    to address this

     if people have to deal with synch with click, then doing
    something like that could be ok

    OP: if multiple clicks on same target, don't have to handle all

    RB: so similar problem then

     we could say calling setptrcapt from within the context of
    gotptrcapt handle isn't legal

    <smaug> just a sec

     If I understand Olli, with click, the spec prevents that (to
    stop recursive issues)

    JR: not sure which is better


      [22] http://www.whatwg.org/specs/web-apps/current-work/#run-synthetic-click-activation-steps

    OP: I kind of like this behavior because it prevents endless

    RB: would be nice if we could say we have the same prob as
    click event
    ... I can look into this

     Seems sensible to make it behave like click

     I can look up the motivation for asynch and add to the bug

    RB: would be good to have a site that uses asynch

    JR: I'll check on the IE bugs that caused this behavior and add
    information to our bug

    <scribe> ACTION: Jacob investigate IE behavior re bug 24971 and
    add that info to the bug [recorded in

    <trackbot> Created ACTION-99 - Investigate ie behavior re bug
    24971 and add that info to the bug [on Jacob Rossi - due

Sub-pixel coordinate granularity

    AB: Rick's March 18 e-mail
    2014JanMar/0190.html> was a followup to a thread he started in
    December 2013
    ... we discussed Rick's December e-mail during our January 7
    m06 and Jacob agreed to add a related non-normative note
    ... however, it appears Rick's March 18 e-mail could be
    touching on different but related issues.

      [24] http://lists.w3.org/Archives/Public/public-pointer-events/2014JanMar/0190.html%3E
      [25] http://lists.w3.org/Archives/Public/public-pointer-events/2013OctDec/0074.html.
      [26] http://www.w3.org/2014/01/07-pointerevents-minutes.html#item06
      [27] https://www.w3.org/2012/pointerevents/track/actions/62.

    RB: the Q about when fractional coords can be returned is an
    open Q

     MSDN has some doc about this

    JR: we are consistent

    RB: do you do that for mouse events too?

    JR: yes, I believe so

    RB: is that causing any issues?

    JR: not positive (we have a CSSOM switch)

    RB: we are going to try to change blink to use float

     and see what sites break

    JR: please let us know the results

    AB: so where are we then?

    RB: I would like to hear more about IE but I don't think our
    spec needs to change

    JR: yes, I agree no PE spec change needed but I can get some
    more info

    <asir> /me my connection keeps dropping today :)

    RESOLUTION: the group agrees there is no need to change the
    spec re the "Sub-pixel coordinate granularity" topic

Touch-action to SVG elements

    AB: Jacob's March 21 e-mail
    2014JanMar/0191.html> was a followup to an e-mail in January
    from Samsung.
    ... Jacob included a proposal in his e-mail. If we agree with
    Jacob's proposal one way forward would be to action Jacob to
    create a bug, submit a changeset and then Resolve/Fix the bug;
    if anyone objects, they can re-open the bug.

      [28] http://lists.w3.org/Archives/Public/public-pointer-events/2014JanMar/0191.html%3E
      [29] http://lists.w3.org/Archives/Public/public-pointer-events/2014JanMar/att-0016/00-part%3E.

    RB: thanks Jacob for the details

     your proposal sounds great to me

    JR: you know Chrome doesn't match it, right?

    RB: yes, and I'll file a bug to get that fixed

     (probably won't get fixed for 35 but will get a fix for 36)

    JR: Olli, same for FF

    OP: yes

    RB: would be nice to have an ED with this fix

    JR: I can do that today

    RB: if have a web site that is broken, that'd be helpful

    AB: so you'll update the spec then Jacob?

    JR: yes, will do

    RESOLUTION: the group agrees Jacob's proposal for "Touch-action
    to SVG elements "topic is OK

Exception usage

    AB: Anne's March 16 followup
    2014JanMar/0185.html> and the original thread starts at
    2014JanMar/0067.html>. Matt has action-73 to followup with Anne

      [30] http://lists.w3.org/Archives/Public/public-pointer-events/2014JanMar/0185.html%3E
      [31] http://lists.w3.org/Archives/Public/public-pointer-events/2014JanMar/0067.html%3E.
      [32] https://www.w3.org/2012/pointerevents/track/actions/73.

    JR: this is a compat problem for IE

     I also am not convinced what we are doing is really a compat

    AB: what about Chrome and FF?

    RB: not sure offhand

     I'd be surprised if this was a problem for us

    OP: I don't think this is an issue for us

    AB: it appears we support the current value

     and we don't want to change it

    JR: if there was strong consensus to change that's one thing
    but I don't think there is

     I feel like we are chasing a moving target

     and only change if there is clear and convincing evidence we
    should change

     Thus I prefer to leave the spec as is

    RB: I agree there doesn't appear to be strong consistency

     so I am ok with leaving spec as is

    AB: DRAFT RESOLUTION: group agrees to keep the spec as is re
    "Exception Usage" topic

    OP: we could ask Anne about the stability of the decision he is

    AB: how about Matt?

    MB: that action-73 is something different


      [33] http://lists.w3.org/Archives/Public/public-pointer-events/2014JanMar/0082.html

    AB: so Matt, Olli do we want to continue to discuss this or
    adopt the Draft Resolution?

    <smaug> smaug annevk: who stable is this DOMException/Error
    handling stuff

      [34] http://lists.w3.org/Archives/Public/public-pointer-events/2014JanMar/0185.html

    MB: Anne pointed to an es_discuss thread that didn't appear to
    have a solid conclusion

    <smaug> annevk smaug: it isn't

     we could see if anything else has been done there in DOM
    and/or WebIDL

     if there is no mature decision about this, I'm OK with
    leaving the spec as is

    AB: would you please followup with Anne then Matt?

    MB: yes, I can do that

    <scribe> ACTION: Brubeck followup with Anne re the Exception
    thread [recorded in

    <trackbot> Created ACTION-100 - Followup with anne re the
    exception thread [on Matt Brubeck - due 2014-04-01].

    SG: for Node, everything is done with codes

    JR: within WebApps, consensus codes is bad

     because it requires centralizing

     and names/strings were supposed to be better in avoiding

    <smaug> annevk smaug: I feel somewhat strongly that new APIs
    should not mint new DOMException names and preferably just
    throw TypeError if there's no branching needs, per Allen's
    preference. and that if people disagree with that, they have a
    discussion with him

    AB: thanks smaug

     one question is if PE is "new" API or not given there are
    sites that are using it

    JR: if there is another UC, that would be helpful

    <jrossi> Issue-65 has been updated (touch-action applies to:)

      [36] https://dvcs.w3.org/hg/pointerevents/raw-file/tip/pointerEvents.html#the-touch-action-css-property

    AB: feels like we should keep this open now

    MB: understand the compat issues for IE

    JR: don't want to change the spec until there is broad
    agreement from the DOM/ES community

     but don't want to change the spec until then

    MB: ok, so sounds like we don't want to change the spec now but
    if DOM or WebIDL spec changes before REC, we could revisit this

    JR: that sounds ok

    AB: sounds ok with me too

     are we back to a resolution we agree not to change spec based
    on what we know now

    JR: yes

    RB: agree

    RESOLUTION: group agrees to keep Exception as is

Feedback on pointer events

    AB: Anne's March 17 followup
    2014JanMar/0187.html>; original thread starts at

      [37] http://lists.w3.org/Archives/Public/public-pointer-events/2014JanMar/0187.html%3E;
      [38] http://lists.w3.org/Archives/Public/public-pointer-events/2014JanMar/0069.html%3E

    JR: I owe Anne a reply

     most is OK

     one part could be contentious re defaultActions if an event
    not canceled

     could add more steps

     but some events are marked canceleable because their mouse
    event counterparts are marked cancelable

     no spec defines that for mouse events

     So I'm not sure we can get to the level of crispness that
    Anne wants

     I think we tried to define this in DOM 3 Events and it was

    DS: my recollection is lots of details is very hard

     I tried with D3E and the feedback was mostly "not good

    AB: so it sounds like Jacob will reply

     and others should join the conversation

Testing: Jacob's latest commits for PR 324 needs review

    AB: Jacob's e-mail
    2014JanMar/0192.html> says PR 324 has been updated. Rick and
    Cathy agreed to review the tests they have previously reviewed.
    ... I don't recall what the means specifically and if that
    means we also need reviews from others.

      [39] http://lists.w3.org/Archives/Public/public-pointer-events/2014JanMar/0192.html%3E

    <smaug> and I lost connection or something

    <smaug> yeah, it is 18:00, I guess Zakim kicked me out

    MB: I split them up

    RB: I think 1-5 are mine

     not sure who reviewed 6-8

    JR: I think the wiki needs some updating

     f.ex. there are some file name updates

    <mbrubeck_> Previously:

      [40] http://www.w3.org/2012/pointerevents/track/actions/45


      [41] http://lists.w3.org/Archives/Public/public-pointer-events/2013OctDec/0029.html

    AB: thanks Matt

     this means review needs to be done by: Rick, Cathy, Art and

    1 will show the diff without whitespace changes.

      [42] https://github.com/w3c/web-platform-tests/pull/324/files?w=1

    RB: were any new files added Jacob?

    <scott_gonzalez> You should be able to add `?w=1` to any diff

    JR: yes, I think so

     check the diff

    <scott_gonzalez> For example,
    it/e872664c81fbcc9d3c53ff5e171ccf48443d066c?w=1 is just the
    most recent commit without whitespace chages.

      [43] https://github.com/InternetExplorer/web-platform-tests/commit/e872664c81fbcc9d3c53ff5e171ccf48443d066c?w=1

    RB: I'll look at anything with "capture"

    JR: could you please check if there are any files without a

    MB: yes, I'll do that

    AB: ok, thanks

    AV: how do we close on the test cases?

    AB: the review should be done by someone that didn't write the

     all comments should be submitted to GH

    JR: I can merge the request after all comments are addressed

    AB: everyone please review your set of tests and put your
    reveiw comments on GH

    JR: if there are big issues, they should be added as GH Issues

    AB: that makes sense to me

CR implementation updates

    AB: any new info re implementations?

    [ None ]


    AB: thanks everything; meeting adjourned

Summary of Action Items

    [NEW] ACTION: Brubeck followup with Anne re the Exception
    thread [recorded in
    [NEW] ACTION: jacob create a changeset for bug 24923 and sent
    it to the list for review [recorded in
    [NEW] ACTION: Jacob investigate IE behavior re bug 24971 and
    add that info to the bug [recorded in

    [End of minutes]

Received on Tuesday, 25 March 2014 16:27:03 UTC