Re: [rtcweb] Video codec selection: Dropping options

On Thu, Nov 28, 2013 at 9:59 AM, Harald Alvestrand <harald@alvestrand.no> wrote:
> On 11/28/2013 11:40 AM, Jeremy Fuller wrote:
>> Since we appear to be entering the realm of voting game theory, +1 to dropping options where consensus has previously proven to be unobtainable. These options have had their moment, time to focus on something else.
>
> The WG has never voted on these alternatives; what failed was an attempt
> at finding consensus.
>
> Furhtermore, the choice of voting method that the chairs are still
> advocating (IRV) is one where presence of non-selected alternatives on
> the slate can influence the outcome.

Without taking a position on which system we should use, I would note
that this is also true of Condorcet.

http://en.wikipedia.org/wiki/Condorcet_method#Evaluation_by_criteria

-Ekr

> I'm not enough of a game theorist to figure out which way the presence
> of the previously non-consensual options may influence the ballot, but I
> don't think I am in agreement with dropping these alternatives from the
> ballot.
>
> (I still haven't decided whether or not I'm in agreement with doing a
> ballot at all, though.)
>
>>
>> "Insanity is doing the same thing over and over and expecting different results"
>>
>> Date: Thu, 28 Nov 2013 12:20:25 +0200
>> From: Leon Geyser <lgeyser@gmail.com>
>> To: "rtcweb@ietf.org" <rtcweb@ietf.org>
>> Subject: Re: [rtcweb] Video codec selection: Dropping options
>> Message-ID:
>>       <CAGgHUiQLXSzU+AvCcvDa383DA=OGd9-NTedfFOAVGt+OmyKwwg@mail.gmail.com>
>> Content-Type: text/plain; charset="iso-8859-1"
>>
>> I also agree on this.
>>
>> Also 5 does not mean anything.
>> I don't like my own option of 8 that much, because 10 is much better :)
>> 11 and 13 are scary options.
>> If 12 is possible it would be a great option. Do most patents only apply for encoding?
>>
>> I am a bit confused about option 15. How does that one work?
>> Let's say I have a two sides: H.264 encoding/decoding + Theora decoder and the other side has VP8 encoding/decoding + Theora decoder. How do they communicate?
>>
>>
>> On 28 November 2013 10:22, <Markus.Isomaki@nokia.com> wrote:
>>
>>> Hi,
>>>
>>> I'm not much in favor of the voting, but if we do it:
>>>
>>> Based on the consensus call we already held in Vancouver, I would
>>> propose to drop from the list in [1] any options that make exclusively
>>> VP8 mandatory to implement or exclusively H.264 mandatory to
>>> implement. I think we have already established that much, and
>>> including those options in any vote seems wrong.
>>>
>>> In practice I'm talking about these four options:
>>> 1. All entities MUST support H.264
>>> 2. All entities MUST support VP8
>>> 3. All entities MUST support both H.264 and VP8 4. Browsers MUST
>>> support both H.264 and VP8, other entities MUST support at least one
>>> of H.264 and VP8
>>>
>>> What the WG should focus on is to see if there is a consensus on any
>>> of the so called "fallback" options vs. no MTI. I think these are all
>>> valid as such as a "fallback":
>>>
>>> 5. All entities MUST support at least one of H.264 and VP8 6. All
>>> entities MUST support H.261 7. There is no MTI video codec 8. 5+6,
>>> i.e. All entities MUST support H.261 and all entities MUST support at
>>> least one of H.264 and VP8 9. All entities MUST support Theora 10. All
>>> entities MUST implement at least two of {VP8, H.264, H.261} 11. All
>>> entities MUST implement at least two of {VP8, H.264, H.263} 12. All
>>> entities MUST support decoding using both H.264 and VP8, and MUST
>>> support encoding using at least one of H.264 or VP8 13. All entities
>>> MUST support H.263 14. All entities MUST implement at least two of
>>> {VP8, H.264 CBP, Theora} 15. All entities MUST support decoding using
>>> Theora.
>>>
>>> I think it would be problematic even if one of options 1-4 came out as
>>> a winner from a voting procedure, since it would force a large part of
>>> the community to implement something they have valid reasons to avoid.
>>> Some of the options 5-15 may have issues as well, but at least a
>>> consensus call among them would still be in order, since we haven't really had it so far.
>>> And presumably the issues and polarization are "smaller" than with the
>>> VP8 vs. H.264 argument.
>>>
>>> Regards,
>>>         Markus
>>>
>>> [1] http://trac.tools.ietf.org/wg/rtcweb/trac/wiki/WikiStart
>>>
>>> _______________________________________________
>>> rtcweb mailing list
>>> rtcweb@ietf.org
>>> https://www.ietf.org/mailman/listinfo/rtcweb
>>>
>> _______________________________________________
>> rtcweb mailing list
>> rtcweb@ietf.org
>> https://www.ietf.org/mailman/listinfo/rtcweb
>
>
> --
> Surveillance is pervasive. Go Dark.
>
> _______________________________________________
> rtcweb mailing list
> rtcweb@ietf.org
> https://www.ietf.org/mailman/listinfo/rtcweb