W3C

- DRAFT -

Touch Events Community Group
27 Jan 2015

Agenda

See also: IRC log

Attendees

Present
Matt_Brubeck, Olli_Pettay, Scott_Gonzalez, Asir_Vedamuthu, Rick_Byers, Tim_Dresser, Mustaq_Ahmed, Doug_Schepers
Regrets
Art_Barstow, Patrick_Lauke
Chair
rbyers
Scribe
tdresser

Contents


rbyers: Agenda should include a discussion of what we should block checkins on for GitHub.

Identifying mouse events derived from touch

<rbyers> https://lists.w3.org/Archives/Public/public-touchevents/2015Jan/0065.html

<rbyers> https://lists.w3.org/Archives/Public/public-touchevents/2015Jan/0078.html

rbyers: Last call, jrossi argued for "FiredFrom" api, Anne had some objections to that Api, as it isn't well defined.
... On the list, I proposed a "sourcedevice" api, which would let you determine if an event came from a device which fires touch events.
... Maybe we should punt until jrossi is here.
... Punting until next call with jrossi, or resolve over email.

Identifying touch events that can trigger scrolling

<rbyers> https://lists.w3.org/Archives/Public/public-touchevents/2015Jan/0058.html

rbyers: We talked about this two weeks ago, came up with a tentative Api, I wrote up a doc for it.
... High level, touch events have a "defaultstartsmanipulation" bit, which indicates if you're about to lose control over touch events.
... It's important that we come up with something that Safari can implement too.

asir: Do we need something like this for other event types as well?

<asir> olli has joined #touchevents

rbyers: We don't have the same kind of state machine with other input types.

olli: It could be useful for pointer events, and wheel events.

rbyers: Chrome's implementation here is really simple, but probably worse than the implementation Firefox has. This isn't necessary for Chrome's current implementation.
... This is somewhat similar to touch action, it could also apply to wheel events.
... jrossi argued that this was more important for touch because touch is common on low power devices, and touch requires lower latency.

Rotation angle / tilt definitions in TEE

<rbyers> https://lists.w3.org/Archives/Public/public-touchevents/2015Jan/0088.html

rbyers: Samsung recently started contributing, which is good to see.
... They believe that they should be firing touch events for stylus, for the 90% case, but they should enable websites expecting stylus input to take advantage of stylus specific features.

shepazu: Wacom is also interested in this.
... Stylus is very important.
... If you use touch for stylus, you'll run into problems with multiple simultaneous modalities (touch and pen at the same time).
... There was an object from a W3C member against publishing pointer events as a formal recommendation.
... They didn't want the fragmentation between touch and pointer events.
... This objection will be raised to the w3c director, but can be delegated. In this case it was delegated to another w3c staff .
... We don't know yet what the official decision will be on that. I expect pointer events will be published as a recommendation.
... One of the points raised in the objection was that there is currently no other way to support pen and touch on the web.
... We could do that here, but I think we want to be careful that we don't just reinvent pointer events and call it touch events.

<rbyers> One possible path (inspired by pointer events) for touch events to address the multi-device scenario: https://docs.google.com/a/chromium.org/document/d/1x-yE4CmTZlZQXea4_FPjgly5TOvF28UBNs9gZ5aGUhg/edit

<smaug> (extending an API has fragmentation issues just like adding a new API)

rbyers: We'll probably want to talk about extending touch events more thoroughly in the future, but for now we do need to respond to Samsung.
... Should we say that adding the stylus specific attributes to touch events are out of scope for now, or should we tell them we'll get back to them in a month?

<mbrubeck> We started adding pen attributes in the abandoned "Touch Events v2" spec https://dvcs.w3.org/hg/webevents/raw-file/default/touchevents.html

shepazu: I would say that it's definitely something that we need to solve for the platform, and long term, we'll consider putting it in touch events.

rbyers: Does anyone object with shepazu's wording?
... I'll reply to the thread then.

Writing

shepazu: Should we have an Api for converting handwriting to text?

<rbyers> http://www.w3.org/TR/ime-api/

rbyers: I don't have much expertise in this space, at Google, our IME team has more context here.

shepazu: Does it seem like this would be useful/possible for the web?

olli: This could be a JS library pretty easily.

rbyers: What's relevant for this group is, "If you want to support this scenario, how would your low-level input apis change?"
... We'd need to expose the historical events. You don't want to optimize for latency in this case, you want to optimize for accuracy.

shepazu: You could do this with a JS library, or an external service. I'll talk to the IME folks.

rbyers: You're right that we do want to enable this offline, you may want to make sure your handwriting isn't being sent across the web.

Process for landing changes in GitHub

<rbyers> https://lists.w3.org/Archives/Public/public-touchevents/2015Jan/0094.html

rbyers: Doug, can you tell us anything about how other groups are using GitHub?

shepazu: Every group is different.
... I'd take pull requests from anyone, and authorized people would integrate the pull requests.
... For editors, the process should be "commit -> review" for everyone else, the process should be "review -> commit"
... The editor can always come to the group and ask for feedback.
... Does that seem reasonable?

rbyers: When you say that for other folks it would be "review->commit", it doesn't need to be a conference call does it? It can just be an editor looking at it?

shepazu: We don't need to look at every little change on the conference call.

rbyers: Here's a proposal:
... If any of us want to land a patch, create a pull request, and then an editor needs to review it.
... Only editors can land patches, and all pull requests will be sent to the list.
... We can always revert changes if there is contention.

<mbrubeck> +1

shepazu: I agree, and think we should accept that.

mbrubeck: I agree, we should keep it light, we can always revert things.

RESOLUTION: use proposed method for landing patches.

shepazu: specs.webplatform.org could pull the spec from GitHub, it will provide a nice way for getting feedback.

rbyers: Right now editors of the extension note are listed as rbyers, sangwhan, shepazu, mbrubeck, is this a good list of editors?

shepazu: Anyone active in the group, making good pull requests, who is a member of the CG should probably be made an editor.
... We should have art on there.
... rbyers should be an editor of v1 errata.

<mbrubeck> rbyers for editor!

rbyers: We want to make sure people like Patrick can easily make a bunch of improvements.

shepazu: Perhaps Patrick should also be an editor.

Plan for next two weeks of calls?

rbyers: I'll be in Australia for the next 2 calls. Should we meet anyways?
... Or we can just meet again in 3 weeks.

shepazu: We might as well cancel.
... rbyers and jrossi are both necessary.

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.140 (CVS log)
$Date: 2015/01/27 20:54:13 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.140  of Date: 2014-11-06 18:16:30  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/asir/olli/
Succeeded: s/??/shepazu/
Succeeded: s/yandex/a W3C member/
Succeeded: s/(Ralph)//
Succeeded: s/your api's/your low-level input apis/

WARNING: Possible internal error: join/leave lines remaining: 
        <asir> olli has joined #touchevents


No ScribeNick specified.  Guessing ScribeNick: tdresser
Inferring Scribes: tdresser
Default Present: Matt_Brubeck, Scott_Gonzalez, rbyers, +1.571.426.aaaa, Doug_Schepers, Olli_Pettay
Present: Matt_Brubeck Olli_Pettay Scott_Gonzalez Asir_Vedamuthu Rick_Byers Tim_Dresser Mustaq_Ahmed Doug_Schepers
Regrets: Art_Barstow Patrick_Lauke
Agenda: https://lists.w3.org/Archives/Public/public-touchevents/2015Jan/0090.html
Got date from IRC log name: 27 Jan 2015
Guessing minutes URL: http://www.w3.org/2015/01/27-touchevents-minutes.html
People with action items: 

WARNING: Possible internal error: join/leave lines remaining: 
        <asir> olli has joined #touchevents



WARNING: Possible internal error: join/leave lines remaining: 
        <asir> olli has joined #touchevents



[End of scribe.perl diagnostic output]