Re: Open issues on Pointer Events - some triage needed

If I recall the last call correctly we went over the issues once and marked
all blocking ones and apparently addressed them all. But it doesn't hurt to
go through them again and make sure we are still okay with addressing the
rest in the next versions or if there is anything we want to fix for v2.

Aside from the spec issues I remember we need 2 passing implementations for
each test to move the spec to recommendation. Patrick, is that still the
case? I took some time to fix a few of our latest bugs and updated the test
results for Chrome 66 here <https://github.com/w3c/test-results/pull/149>.
There are still some tests
<https://rawgit.com/NavidZ/test-results/afd4c908535310e1b1ed422e184ffe1742ad72a2/pointerevents/less-than-2.html>
 that have less than 2 implementation passing.


   - */pointerevents/extension/**: we can ignore these as they will be for
   V3 of pointerevents or something like that
   - *tangentialPressure* and *twist* are the fields that we added in
   Chrome but based on the test results they are not still in FF and Edge.
   - touch-action values like *pan-** are added in Chrome but not in FF or
   Edge based on the test results.
   - */pointerevents/pointerevent_click_during_capture-manual.html:* This
   was something we agreed on some behavior in this issue
   <https://github.com/w3c/pointerevents/issues/75>. But when I changed
   Chrome to that there were some websites not behaving correctly. There is no
   explicit text around this in pointerevents and the behavior we agreed on
   was just a deduction based on ui events and pointerevents interaction. Also
   different browsers apparently do different things here. I suggest we can
   skip this test and the corresponding issue to investigate the compat issues
   over time. What do you think?


And most of the other tests I believe are missing the results from other
vendors assuming I haven't missed anything else.

Olli, what do you think about the tangentalPressure and twist as well as
pan-* touch actions in FF? Will you be able to add them to FF or should we
push them to the extension? Can you get the latest results for FF in that
repository so it is easier to catch what's missing there and maybe discuss
it in the next voice call.

Scott Lo, how about Edge? Are there any plans for adding
tangentialPressure/twist/pan-* to pointer events? Can you also get the Edge
pointerevent test results updated?

Patrick, can we go with Olli's suggestion and setup the first voice call?
We used to have them on Wednesday mornings if I recall correctly? Is that
still good for people?

Cheers,
Navid


On Mon, Feb 12, 2018 at 2:07 PM, Olli Pettay <olli@pettay.fi> wrote:

> On 02/12/2018 05:34 PM, Patrick H. Lauke wrote:
>
>> Dear all,
>>
>> as discussed, after some inactivity, I'm hoping to get back into gear and
>> for the group to push Pointer Events 2 out, as for the most part I believe
>> we're in a solid position with some good actual real-world browser support.
>> However, there are quite a few issues still open against the spec - in many
>> cases, there was initial discussion and activity, but it then petered out.
>>
>> I've gone through and tried to put all issues (51, at the time of
>> writing) into some rough "buckets":
>>
>> * TESTS https://github.com/w3c/pointerevents/issues?q=is%3Aissue+is%
>> 3Aopen+label%3Atest
>> these appear to be mostly related to test framework?
>>
>> * ENHANCEMENT https://github.com/w3c/pointer
>> events/issues?q=is%3Aissue+is%3Aopen+label%3Aenhancement
>> these feel to me like nice-to-have/feature requests that go beyond what
>> we're currently aiming for in PE2 (or that would take quite a bit of time
>> to flesh out properly at this point)
>>
>> * QUESTION https://github.com/w3c/pointerevents/issues?q=is%3Aissue+is%
>> 3Aopen+label%3Aquestion
>> a mixed bag of questions/clarifications - these are probably the ones we
>> should address most urgently if they relate to anything being vague/unclear
>>
>> *BUG https://github.com/w3c/pointerevents/issues?q=is%3Aissue+is%
>> 3Aopen+label%3Abug
>> probably the wrong term, but used this for things that are arguably a bug
>> in our spec text (i.e. a clear omission)
>>
>> Now my question: would it be possible for all of us collectively to have
>> a look through these issues and to triage them? Keeping in mind we'd
>> ideally want to get to a point very soon where everything in scope for PE
>> is addressed (with sufficient clarity in spec language), and anything that
>> is a new/nice to have/future is deferred for possible PE Level 3 (whether
>> this happens is still open to discussion).
>>
>> With a lot of these issues, the material goes quite outside of my
>> technical comfort zone, but if you can help me out with a rough outline of
>> what the spec may need to say / what needs changing or clarifying, I can
>> help out with further wordsmithing.
>>
>> Do WG members feel it would be good to fall back to regular weekly voice
>> calls? Or are we still happy working/coordinating via GitHub?
>>
>
>
> Triaging calls can be rather effective to sort out what to do with (spec)
> issues. Maybe not every week, but twice a month? And once all the issues
> have been gone through, just stop having calls again.
>
> -Olli
>
>
>> Let me know,
>>
>> P
>>
>
>
>