Re: [rtcweb] VP8 vs H.264 - the core issue

["cb.list6" <[e-mail]>]

On Oct 25, 2013 4:32 PM, "Justin Uberti" <[e-mail]> wrote:
>
>
>
>
> On Fri, Oct 25, 2013 at 2:47 PM, cowwoc <[e-mail]> wrote:
>>
>> On 25/10/2013 3:50 PM, Justin Uberti wrote:
>>>
>>> No, this is a fundamental issue. If H.264 is the MTI, to quote Harald,
"we give up and live with royalties forever". This will have a significant
impact on WebRTC innovation and adoption.
>>>
>>> The fact that there are some who disagree with the IPR status of VP8
does not invalidate this point.
>>
>>
>>     What prevents us from retiring H264, VP8 or any other codec in the
future? Surely we are allowed to retire codecs after 5-10 years? If not,
listing any codec as MTI (including VP8) places us at a some level of risk.
>>
>>
>>> Nobody is saying implementors can't support H.264. If interoperability
with such services is a high priority, implementors will ship H.264
regardless of what we decide here.
>>
>>
>>     The same is true for VP8, so why mandate it as required?
>
>
> Because we want to guarantee video interoperability without forcing
everyone to pay licensing feeds.
>

That's a fine goal. But there are camps in this wg that want license fees
paid.

So there will be no concensus.

>>
>> If VP8 is as good as everyone alleges (and I think it is) then surely
vendors should be moving in this direction. But they are not. Forcing VP8
as MTI comes across as a political ploy to force their hand.
>
>
> Regarding VP8, pretty much every SoC vendor is now shipping HW VP8
acceleration, so I think there is clear movement in this direction.
>>

Ack. Then the market may favor vp8.

>>
>>     From an end-user point of view, the only thing I really care about
is getting WebRTC support from Microsoft and Apple. Because I view
everything through that prism, my primary concern is how this decision
affects them jumping on board.
>
>
> I can't speak for Microsoft and Apple, but I don't think they are sitting
around waiting for us to make up our mind on this.
>>

Right.  So it sounds like you agree there is no need for mti since each
implementation will drive it's own strategy.

CB

>>
>> Gili
>>
>> _______________________________________________
>> rtcweb mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/rtcweb
>>
>
>
> _______________________________________________
> rtcweb mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/rtcweb
>