Web Real-Time Communications Working Group Teleconference
14 Jun 2011


See also: IRC log


Caroline, StefanH, Salvatore, nstratford, Tim, Matt, Alissa, Dan, hta, Druta, Cullen, Rian, Wu_Chou
Francois, Dan_Burnett
Harald, Stefan


<hta> Can someone fix the topic line? It's 1600Z, not 1700Z...

<trackbot> Date: 14 June 2011

<hta> "the conference is restricted at this time". Only open after 1600Z?

<Salvatore> is it possible joining using SIP ?

<hta> Technical issues. I'm getting a coffee while we wait.

<StefanH> anyone getting into the conf bridge?

<anant> nope, still says restricted

<Salvatore> the same for me

<StefanH> very strange; according to http://www.w3.org/Guide/1998/08/teleconference-calendar#s_4756 it should be on now

<Wu> We cannot get on to the conference bridge. What is the correct conference code 78294?

<StefanH> @Wu: should work now

<hta> Wu, yes it is. what error did you get?

<anant> Wu: yes, followed by #

<hta> rian

<scribe> scribe: Matt

<scribe> Agenda: http://lists.w3.org/Archives/Public/public-webrtc/2011Jun/0005

IETF meeting

hta: IETF is trying to get focus. Getting clarity on some things, and not on others.
... We know we'll be managing RTP data is for sure, how is not so clear.
... Security requirements discussion, how much is needed and how to apply it.
... Congestion control is a hot topic, IETF says it's not optional to go without.

cullenfluffyjenn1: One comment on the notes: the comment that the datagram transport looks like DTLS, was the direction people were going. I took that differently as one of the possibilities.
... I don't think there is agreement on that.

hta: I can't see anyone proposing anything else in that context.

cullenfluffyjenn1: Some are proposing DTLS over DCCP over UDP, which has application layer congestion control.

hta: Other questions about IETF meeting?

<cullenfluffyjenn1> Some people proposing DTLS/UDP some DTLS/DCCP/UDP

<derf> Yes.

Proposal to adopt API as one starting point

hta: I think this will be the main discussion. The Google/WHATWG API has been written up and is being discussed.

matt: Was that one starting point or a starting point?

hta: English translation maybe it is not "THE starting point". We don't want to do anything that would preclude anyone coming up with another proposal.

Wu: Can you post a link to the proposed starting point?

<Wu> Wu

<StefanH> http://www.whatwg.org/specs/web-apps/current-work/webrtc.html

-> http://www.whatwg.org/specs/web-apps/current-work/webrtc.html WebRTC API from WHATWG

cullenfluffyjenn1: I am a little confused. If we adopted this, what would you want people to provide? Changes? Or alternatives?

hta: My preferred way to go forward would be both approaches. If people think this is something that is useful they propose changes, and if the feedback is that it isn't any good, that we work on an alternative.
... If there are multiple sets of requirements and this fulfills one of them, and then we may need additional APIs, which others can share.

cullenfluffyjenn1: I'd like to see more discussion on this doc, or other alternatives before making this decision.
... I imagine we could change this document to meet our needs, but given that there's been zero talk about it, I think it's too early to make that decision.

StefanH: One of the reasons to have this, is because the discussion is going on outside of the WG. Otherwise there would be parallel discussion elsewhere.

cullenfluffyjenn1: Is the author on the call? Can they move it to this group?

hta: The author has promised to take input, but not read all mailing lists.

Wu: I would suggest we establish a set of applications that the spec intends to support.

<hta> Matt, the author is Ian Hickson.

Wu: Then we examine these APIs and decide if these are sufficient, and if not how we can improve the API. I think we should establish requirements first before locking into an API.

StefanH: I agree, we should not lock into one API at this stage. This proposal is just one proposal.
... We encourage other alternatives.

cullenfluffyjenn1: How do you want to get alternative proposals?

Wu: Our plan should be to establish the supported applications and requirements. Look at the potential APIs and look at whether they support those applications. e.g. look and see if Google's API supports our application requirements.

<alissa> will write in the chat

<alissa> In DAP we took inputs from individual members

<alissa> e.g., http://lists.w3.org/Archives/Public/public-device-apis/2009Apr/att-0001/calendar.html

<alissa> rather than making every API a working draft of the group

<alissa> although we also have some working drafts that will probably never go to rec as well...just offering it as a suggestion

hta: I don't know if that is reasonable, or if we're using language wrong. We'd like to have the document as an official input into the WG, that is clearly contributed to the WG.

<alissa> in my experience "working draft" in the W3C and "WG draft" in the IETF are not exactly the same thing

Wu: One process would be to invite Member Submissions.
... I would treat Google API as a Member Submission.

hta: This API does not belong to Google, it claims no rights to it or it's contents. The Editor works for Google but that is all.

Wu: Someone should submit it, it then does not represent the position of the WG. A WD is arrived at from consensus in the WG.

cullenfluffyjenn1: Plus one.

<alissa> who would the member submission come from if it's not written by a member (e.g., if whatwg is not a W3C member)?

hta: Is it a consensus position from the WG that the chairs should solicit input from the members for proposals and about the WHATWG proposal?

<alissa> (i.e., if Google is not submitting it)

cullenfluffyjenn1: Yes, but who would the submission come from. Which member brings it in, or how W3C handles this...

Wu: I think the typical practice is that members or multiple members can jointly or individually can submit input to a WG, and then the WG can work on it, but it does not just become the WD. WD are based on consensus of the WG.

<alissa> agreed on submissions being able to come from multiple members. DAP had contributions from BONDI, don't think BONDI was a member (it's a consortium)

Wu: From example the ?? was submitted by IBM and others. The final WD had to be drafted by the WG and get WG consensus.

hta: The chairs will create a call for submissions and send to the mailing list.

anant: Is there another proposal beyond the WHATWG API?

hta: If so, I am unaware.

<anant> no worries!

<derf> Phones are hard.

<derf> But that's the subject of a different working group.

<scribe> ACTION: Harald will prepare a call for contributions and make sure the WHATWG draft is contributed [recorded in http://www.w3.org/2011/06/14-webrtc-minutes.html#action01]

<trackbot> Created ACTION-1 - will prepare a call for contributions and make sure the WHATWG draft is contributed [on Harald Alvestrand - due 2011-06-21].

API requirements

Wu: I would recommend that you post the link of these requirement documents for the record please.

<cullenfluffyjenn1> I'll post the link - just a sec

Wu: The minutes will be public, so we do this so the public can view what we discussed.

<cullenfluffyjenn1> http://tools.ietf.org/html/draft-holmberg-rtcweb-ucreqs-01

-> http://tools.ietf.org/html/draft-holmberg-rtcweb-ucreqs-01 Web Real-Time Communication Use-cases and Requirements

hta: The use cases themselves have a good reason to live in a place in common between IETF and W3C.
... The requirements on the API would probably have a better home in a W3C WD.

matt: Sounds right to me.

hta: To make this happen, we need a draft and an editor.

StefanH: Göran is not on the call, but he has promised that he can edit this document.

hta: Sounds like a reasonable start.

<scribe> ACTION: Göran to propose a first draft of requirements document [recorded in http://www.w3.org/2011/06/14-webrtc-minutes.html#action02]

<trackbot> Created ACTION-2 - Propose a first draft of requirements document [on Göran Eriksson - due 2011-06-21].


<trackbot> ACTION-2 -- Göran Eriksson to propose a first draft of requirements document -- due 2011-06-21 -- OPEN

<trackbot> http://www.w3.org/2011/04/webrtc/track/actions/2

hta: I bet StefanH will ask for a reasonable date from Goran and we'll get back to it.

F2F meetings

hta: We don't have certainty on when we'll have the meeting in Quebec City, but either Saturday or Sunday before the IETF meeting.

cullenfluffyjenn1: With my IETF hat on, Gonzalo from the IESG is handling this right now. I know he's pursuing getting us a room for Saturday, but didn't have one yet.

hta: Would be great to have actual documents at the F2F and have technical discussions there. Seems like we can get there.

TPAC dates: 31 October to 4 November 2011

-> http://www.w3.org/2011/11/TPAC/ TPAC agenda

hta: Francois is trying to schedule us for Monday and Tuesday of that week.

DanR: Did Gonzalo mention if there would be network running?

cullenfluffyjenn1: We didn't talk about that. For some other meetings, the code sprint, the Internet of things hackfest, etc, are expecting to have network. I hope we will, but didn't discuss it.

hta: We'll have to come back on the mailing list for discussion about the agenda and so on.

Agreement with DAP on streams generation

<Dan> that was Dan (R) not Wu

StefanH: DAP has camera API for files and images, while we would handle streams.

hta: No comments? Then we will do that going forward.
... And what about the inverse? Do we need to have some of our requirements turned back into DAP for control of things like microphone settings.
... Before we get the requirements down it is hard to tell if we will have to do that, but it is a possibility.
... We'll take that discussion to the list too.
... More reasonable to discuss that after requirements.

matt: Is anyone in RTC also in DAP?

hta: I'm on the mailing list.

<alissa> I'm a member of DAP but not very active.

matt: I'm just thinking you might want someone who is the eyes and ears of RTC for DAP, so you can see when things are going to overlap, or drop, or whatever.

hta: We could make an action item for someone to be the DAP/RTC liaison.

<scribe> ACTION: StefanH to find someone to monitor DAP for RTC. [recorded in http://www.w3.org/2011/06/14-webrtc-minutes.html#action03]

<trackbot> Sorry, couldn't find user - StefanH

<scribe> ACTION: Stefan to find someone to monitor DAP for RTC. [recorded in http://www.w3.org/2011/06/14-webrtc-minutes.html#action04]

<trackbot> Created ACTION-3 - Find someone to monitor DAP for RTC. [on Stefan Håkansson - due 2011-06-21].

Any other business?

hta: Anything?

cullenfluffyjenn1: I have a technical question that I'd love to get input from people.
... Do we think that the API needs to be able to expose up what the codec capabilities are up to the JavaScript level?
... One way would be to let the lower layer deal with it, another would be to let the API sort it out. Have people thought about this? Strong feelings?

StefanH: There should be a way to negotiate the codec, but that doesn't mean it has to be exposed in JavaScript.

anant: I think it's important for JavaScript to expose capabilities to webapps. Take the teleconference scenario, the negotiation needs to happen at the app layer to figure out what codec should be used.

hta: As a contributor: I've had pushback in the past saying that the API needs to be designed so that it is possible to use it without caring about those details.
... The WHATWG proposal separates these a bit into stuff that is opaque and stuff that the app needs to know
... The codec information is hidden in the opaque blobs of data.
... The need to be able to do an application without caring about codecs and still get them correctly negotiated is a requirement for people I talked to.

anant?: We still need cross platform work. If we have codecs that don't work, we need the app to know that.

Tim: There are also app specific modes that an app may want to support, e.g. a music mode and talk? mode. Which one you use depends on the application. While you may not care which one is used, but if you could it'd be better.

cullenfluffyjenn1: I like that you can do it without worrying, but it is important to be able to understand what is going on when there is a failure. Parameterize speech vs music, big video vs little, etc.

anant: I am for specifying sane defaults: an app doesn't have to care about codecs. But we need to expose it, whether all apps use it or not.

StefanH: I'd like to bring this into the recent discussion of ICE. I think a use case is to do this without having to implement ICE functionality.

hta: Of the people who have strong opinions on what applications can be and need: please think about whether you can contribute text for a use case.
... That way we capture the use cases in which features are needed as we go forward.

matt: Where should text go? ML or wiki, or...

hta: If the contributor is active in IETF, send it to their mailing list, if not webrtc mailing list. Someone there can carry it forward.

-> http://www.w3.org/2011/06/14-webrtc-minutes.html#ActionSummary Action Items

StefanH: Should we have another call before IETF?

hta: Good question, but you'll have to handle it as I am on holiday the next three weeks.
... Want another call?

cullenfluffyjenn1: Yes, calls help make progress.

hta: You're volunteering then!

cullenfluffyjenn1: Will it maybe make sense to have a call three weeks before the IETF? I think we should have one well before the meeting at the IETF, so that everyone can prep, etc.

matt: The calls also serve as a good heartbeat. They let you know whether you've made progress or not every single week.

hta: If we go five weeks that's 5 July.

cullenfluffyjenn1: I think 12 July is better, given other deadlines

hta: I'll be back for 12 July.

StefanH: I'll be gone that week.

<hta> matt, I'll be back July 12, not July 5

StefanH: Let's do a Doodle form.

<scribe> ACTION: Harald to create a Doodle poll for next call meeting date. [recorded in http://www.w3.org/2011/06/14-webrtc-minutes.html#action05]

<trackbot> Created ACTION-4 - Create a Doodle poll for next call meeting date. [on Harald Alvestrand - due 2011-06-21].

hta: Thank you all!

<cullenfluffyjenn1> bye

Summary of Action Items

[NEW] ACTION: Göran to propose a first draft of requirements document [recorded in http://www.w3.org/2011/06/14-webrtc-minutes.html#action02]
[NEW] ACTION: Harald will prepare a call for contributions and make sure the WHATWG draft is contributed [recorded in http://www.w3.org/2011/06/14-webrtc-minutes.html#action01]
[NEW] ACTION: Harald to create a Doodle poll for next call meeting date. [recorded in http://www.w3.org/2011/06/14-webrtc-minutes.html#action05]
[NEW] ACTION: Stefan to find someone to monitor DAP for RTC. [recorded in http://www.w3.org/2011/06/14-webrtc-minutes.html#action04]
[NEW] ACTION: StefanH to find someone to monitor DAP for RTC. [recorded in http://www.w3.org/2011/06/14-webrtc-minutes.html#action03]
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.136 (CVS log)
$Date: 2011/06/14 17:00:02 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.136  of Date: 2011/05/12 12:01:43  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/@@/DTLS over UDP/
Succeeded: s/@@/DTLS/
Succeeded: s/DTLS over UDP/DTLS over DCCP over UDP/
Succeeded: s/??/Wu/
Succeeded: s/@@/both approaches/
Succeeded: s/??/Göran/
Succeeded: s/persuing/pursuing/
Succeeded: s/ISEG/IESG/
Succeeded: s/Wu:/DanR:/
Succeeded: s/??/StefanH/
Succeeded: s/??/Tim/
Succeeded: s/use ICE/implement ICE/
Succeeded: s/it's/the contributor is active in/
Succeeded: s/for 5 July/for 12 July/
Found Scribe: Matt
Inferring ScribeNick: matt
Default Present: Caroline, Ralph, +, +358.408.20aabb, Matt, Alissa, +972.3.645.aacc, +1.404.978.aadd, +1.425.391.aaee, nstratford, +1.403.244.aaff, hta, StefanH, Salvatore, Dan, Cullen, +44.203.014.aagg, Druta, Rian, Tim, Anan, Wu_Chou, anant
Present: Caroline StefanH Salvatore nstratford Tim Matt Alissa Dan hta Druta Cullen Rian Wu_Chou
Regrets: Francois Dan_Burnett
Agenda: http://lists.w3.org/Archives/Public/public-webrtc/2011Jun/0005
Found Date: 14 Jun 2011
Guessing minutes URL: http://www.w3.org/2011/06/14-webrtc-minutes.html
People with action items: g harald ran stefan stefanh

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.

[End of scribe.perl diagnostic output]