Re: [rtcweb] ~"I'd love it if patents evaporated...If not now, when"

[cowwoc <[e-mail]>]

On 13/11/2013 7:36 PM, Ron wrote:
> On Wed, Nov 13, 2013 at 05:32:17PM -0500, cowwoc wrote:
>> Start with H.261 and replace it the moment you find something better.
> Except this would essentially be an RFC 6919 MUST (BUT WE KNOW YOU WON"T)
> recommendation, since nobody is going to spend development effort on a
> thing that gives thumbnail video that wasn't even acceptable for users
> in the 80's.

How realistic is that statement? Today I think I can get 720p (or was it 
1080p, I forget) using VP8 in 1.5Mbps.

How much bandwidth would you need to do the same in H.261? What 
resolution would get you for 1.5Mbps? We need some reasonable comparisons.

>> Forcing us to transcode or drop video calls is not a solution.
> ... which means you'll still have exactly this problem - no solution.
>
> If MTI video is important, the selected codec does need to be better
> than the equivalent of mandating morse as the MTI audio channel.
>
>
> Either we mandate something that is of good quality and has a licence
> that allows anyone to use it - or we accept that the people trying to
> hijack the potential ubiquity of this standard have succeeded in that.
>
> Seems like a pretty simple choice to me.

I think you're exaggerating how bad H.261 is and it's also misleading to 
imply that the vast majority of users would be forced to use H.261. We 
are talking about H.261 as a fallback (if upgrading to VP8 or H.264 
fails). In the 5-10% of cases this happens, I am advocating we get using 
H.261 instead of dropping the call.

Gili