Video from last week's country meetings.

C. Scott Ananian cscott at
Tue May 27 09:35:11 EDT 2008

On 5/27/08, NoiseEHC <NoiseEHC at> wrote:
> If you have missed:

Traditionally the problem with Ogg for video conferencing is
*encoding*, not decoding.  General purpose video codecs are usually
designed to be asymmetric in that way: very cheap to decode, but
sometimes expensive to encode (depending on how much compression you

Talking with Rob Savoy at FISL, he mentioned that recently some
speedups for Ogg encoding has been found that offered over an order of
magnitude improvement.  I can't tell if that's what's being discussed
in the email above, or if these improvements are still in the pipe.
But Rob promised that we'd have a full end-to-end Ogg solution by the
end of the year.

See also
-- although that uses the patent-encumbered gsm codecs, I believe.
 ( search for 'patent')

On the "trolling for devel help" angle: the linked message from
Guillaume mentioned sugar integration issues.  Indeed, we'd like to
unify the API/UI used by chat and "telephone calls" with the
invitation API/UI, so that inviting a friend for a game of chess looks
& feels (and is implemented) the same as receiving a telephone call
from the friend.  If you're interesting on hacking on this, let me

                         ( )

More information about the Devel mailing list