Re: Call for Consensus (CfC): Publish webrtc-stats as updated Candidate Recommendation

If it is a matter of weeks, I would agree. But if it requires review and discussion at TPAC (and subsequent editing) so that it is more like 3+ months then maybe not.

Henrik - what do you think?

On Jul 16, 2019, at 11:12 AM, Stefan Håkansson LK <stefan.lk.hakansson@ericsson.com<mailto:stefan.lk.hakansson@ericsson.com>> wrote:

I’m not heavily involved in this anymore. But to me it looks like moving the (per-layer) stats to the correct place, and decide & write up what to do with those meaningless to aggregate should be a matter of weeks. Perhaps it would be better to do that before issuing a new CR (since the move would anyway happen soon-ish I presume, and then a CR issued now would be obsolete even though being relatively fresh). My 5 cents.

From: Harald Alvestrand <harald@alvestrand.no<mailto:harald@alvestrand.no>>
Date: Tuesday, 16 July 2019 at 20:00
To: Henrik Boström <hbos@google.com<mailto:hbos@google.com>>, Bernard Aboba <Bernard.Aboba@microsoft.com<mailto:Bernard.Aboba@microsoft.com>>
Cc: Stefan LK <stefan.lk.hakansson@ericsson.com<mailto:stefan.lk.hakansson@ericsson.com>>, "public-webrtc@w3.org<mailto:public-webrtc@w3.org>" <public-webrtc@w3.org<mailto:public-webrtc@w3.org>>
Subject: Re: Call for Consensus (CfC): Publish webrtc-stats as updated Candidate Recommendation

When we (the chairs) discussed issuing a renewed CR, it was explicitly clear that this is not stable enough for PR yet.
We'll likely be issuing another CR before asking for PR.

When we discussed the per-layer stats, I felt that there was value in having the (current) aggregated stats, and augumenting them with
per-outbound-rtp stats; the exception is those that are really meaningless to aggregate, such as framerate and width/height; those either need to be deprecated (moved to the obsolete section) at the per-sender level or need to be declared as "identical to the highest level sent".

On 7/16/19 6:08 PM, Henrik Boström wrote:
OK, thanks for the clarification. The current state of the spec is in a better state than the currently published version though.

On Tue, Jul 16, 2019 at 5:35 PM Bernard Aboba <Bernard.Aboba@microsoft.com<mailto:Bernard.Aboba@microsoft.com>> wrote:
Henrik said:

"several metrics are in the wrong place, i.e. to adequately support simulcast some of them need to be moved around so that they can be expressed on a per-layer (per outbound-rtp) rather than per-sender (track) basis."

[BA] Candidate Recommendation is supposed to denote "ready for implementation". If we have metrics in the wrong place, wouldn't it be better to move them around now so that when we put out the CR, we can then say "this is how you support stats for simulcast" and move forward on implementation?

Stefan said:

"But should the “CR blocker” labels be changed to “PR blocker”? "

[BA]  The metric move issue above does indeed seem like a "CR blocker".   If there's nothing else that would preclude an implementation push, perhaps we can just remove the "CR blocker" label on other issues that aren't critical for issuing a CR.



________________________________
From: Henrik Boström <hbos@google.com<mailto:hbos@google.com>>
Sent: Tuesday, July 16, 2019 1:31 AM
To: Stefan Håkansson LK <stefan.lk.hakansson@ericsson.com<mailto:stefan.lk.hakansson@ericsson.com>>
Cc: Bernard Aboba <Bernard.Aboba@microsoft.com<mailto:Bernard.Aboba@microsoft.com>>; Harald Alvestrand <harald@alvestrand.no<mailto:harald@alvestrand.no>>; public-webrtc@w3.org<mailto:public-webrtc@w3.org> <public-webrtc@w3.org<mailto:public-webrtc@w3.org>>
Subject: Re: Call for Consensus (CfC): Publish webrtc-stats as updated Candidate Recommendation

I don't know the formal differences between CR and PR. But in my mind, the stats are in a really good shape - the desired metrics for a WebRTC 1.0 implementation are there, and some - but several metrics are in the wrong place, i.e. to adequately support simulcast some of them need to be moved around so that they can be expressed on a per-layer (per outbound-rtp) rather than per-sender (track) basis.

On Tue, Jul 16, 2019 at 10:17 AM Stefan Håkansson LK <stefan.lk.hakansson@ericsson.com<mailto:stefan.lk.hakansson@ericsson.com>> wrote:

I do too. But should the “CR blocker” labels be changed to “PR blocker”? The current label seems to serve no purpose if we spin a new CR without resolving the issues.



From: Henrik Boström <hbos@google.com<mailto:hbos@google.com>>
Date: Monday, 15 July 2019 at 13:24
To: Bernard Aboba <Bernard.Aboba@microsoft.com<mailto:Bernard.Aboba@microsoft.com>>
Cc: Harald Alvestrand <harald@alvestrand.no<mailto:harald@alvestrand.no>>, "public-webrtc@w3.org<mailto:public-webrtc@w3.org>" <public-webrtc@w3.org<mailto:public-webrtc@w3.org>>
Subject: Re: Call for Consensus (CfC): Publish webrtc-stats as updated Candidate Recommendation
Resent-From: "public-webrtc@w3.org<mailto:public-webrtc@w3.org>" <public-webrtc@w3.org<mailto:public-webrtc@w3.org>>
Resent-Date: Monday, 15 July 2019 at 13:23



I support publishing the current webrtc-stats as CR





On Fri, Jul 12, 2019 at 8:29 PM Bernard Aboba <Bernard.Aboba@microsoft.com<mailto:Bernard.Aboba@microsoft.com>> wrote:

The respec errors in Webrtc-Stats are now fixed:

https://w3c.github.io/webrtc-stats/<https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fw3c.github.io%2Fwebrtc-stats%2F&data=02%7C01%7CBernard.Aboba%40microsoft.com%7Cfc3461a0b9ff416f194108d70a191fa6%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636988975338841511&sdata=xNZ0rxiWEGPS5dniOqYF6Qaq6CO36hWlMJ5I%2BUBRmII%3D&reserved=0>

________________________________

From: Bernard Aboba <Bernard.Aboba@microsoft.com<mailto:Bernard.Aboba@microsoft.com>>
Sent: Thursday, July 11, 2019 8:09 AM
To: Harald Alvestrand <harald@alvestrand.no<mailto:harald@alvestrand.no>>; public-webrtc@w3.org<mailto:public-webrtc@w3.org> <public-webrtc@w3.org<mailto:public-webrtc@w3.org>>
Subject: Re: Call for Consensus (CfC): Publish webrtc-stats as updated Candidate Recommendation



The Editor's draft<https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fw3c.github.io%2Fwebrtc-stats%2F&data=02%7C01%7CBernard.Aboba%40microsoft.com%7Cfc3461a0b9ff416f194108d70a191fa6%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636988975338841511&sdata=xNZ0rxiWEGPS5dniOqYF6Qaq6CO36hWlMJ5I%2BUBRmII%3D&reserved=0> shows 11 respec errors, and the Issues list<https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fw3c%2Fwebrtc-stats%2Fissues&data=02%7C01%7CBernard.Aboba%40microsoft.com%7Cfc3461a0b9ff416f194108d70a191fa6%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636988975338851508&sdata=ylaiKag51vWMt4ANDEo3igqB9STorXR6If5IobddFDE%3D&reserved=0> shows 8 unresolved Issues with the "CR blocker" label.



So perhaps a little more work is needed before a CR can be published.

________________________________

From: Harald Alvestrand <harald@alvestrand.no<mailto:harald@alvestrand.no>>
Sent: Wednesday, July 10, 2019 7:52 AM
To: public-webrtc@w3.org<mailto:public-webrtc@w3.org> <public-webrtc@w3.org<mailto:public-webrtc@w3.org>>
Subject: Call for Consensus (CfC): Publish webrtc-stats as updated Candidate Recommendation



Our present webrtc-stats, published as https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.w3.org%2FTR%2Fwebrtc-stats%2F&amp;data=02%7C01%7CBernard.Aboba%40microsoft.com%7C0d5b18e330f34e01f29c08d705467791%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636983672521933649&amp;sdata=efffKD9Wp7Rd0D9Vyxvb5jAHeY0xiwBINk1xxdE6WQY%3D&amp;reserved=0<https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.w3.org%2FTR%2Fwebrtc-stats%2F&data=02%7C01%7CBernard.Aboba%40microsoft.com%7Cfc3461a0b9ff416f194108d70a191fa6%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636988975338851508&sdata=CwTb8y9sxtr7xPoFCPjkpsu8mVfjfJYemnoAgZS%2BfHo%3D&reserved=0>, is almost exactly 1 year old.

The spec has evolved significantly over the course of a year; having this old version
around just confuses things.

The spec is still evolving, and needs some tidying, but the chairs feel that it is better to emit this one
as Candidate Recommendation now than to wait for stability to increase. Thus:

This is a Call for Consensus to push the editors' draft,
https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fw3c.github.io%2Fwebrtc-stats%2F&amp;data=02%7C01%7CBernard.Aboba%40microsoft.com%7C0d5b18e330f34e01f29c08d705467791%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636983672521933649&amp;sdata=fcxkyzDASKpqrkZMrozRSh8HxSLy34wVW8z3BSSO6xk%3D&amp;reserved=0<https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fw3c.github.io%2Fwebrtc-stats%2F&data=02%7C01%7CBernard.Aboba%40microsoft.com%7Cfc3461a0b9ff416f194108d70a191fa6%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636988975338861504&sdata=Lw1IJFoPwknCCK9jezaOdAdTi8hnZ3zULc0q6AdkgIM%3D&reserved=0>, out as a new version of our
Candidate Recommendation.

The CfC will last for one week, and end on Wednesday, July 17, at 16:00 UTC

In response, please state one of the following:

- I support publishing the current webrtc-stats as CR

- I object to publishing the current webrtc-stats as CR, due to
issues filed in open bug <#number>.

Harald, for the chairs

--
Surveillance is pervasive. Go Dark.




--

Surveillance is pervasive. Go Dark.