Cerowrt RC6 plan

Cerowrt RC6 plan

[Dave Taht]

As a core goal of cerowrt is to stay no more than 2 release cycles
behind the kernel mainline, and we're running several weeks behind the
original release date for cerowrt, anyway, updating the kernel past
2.6.39 has been on my mind.

So, thx to the hard work of 'solca', I just got a build of linux 3.0.3
done which thus far appears stable. The only blocker for RC6 that I
have is:

http://www.bufferbloat.net/issues/243

And the problem with ipv6 and TOS irks me. I keep hoping I'm merely
doing something silly, but more eyeballs on:

http://www.bufferbloat.net/issues/243#change-870

would be nice.

So... I hope to get RC6, containing the 3.0.3 kernel and fixed
iptables out by saturday, as I am finished with linuxcon and will be
back in isc's lab by friday. Proving that stable as a base platform
will enable RC7 to be mostly fixes and additions to various utilities.

Numerous other small fixes have gone in as well,

http://www.bufferbloat.net/projects/cerowrt/issues?query_id=5

but many of the more difficult ones - notably gui-related bugs - seem
likely to be unfixable by 1.0. I do hope we can address some more of
the time and dnssec related bugs by RC7, which is tenatively targetted
for sept 5.

http://www.bufferbloat.net/projects/cerowrt/issues?query_id=2


-- 
Dave Täht
SKYPE: davetaht
US Tel: 1-239-829-5608
http://the-edge.blogspot.com

Re: Cerowrt RC6 plan

[Dave Taht]

oops, the thing I was hoping for more eyeballs on was:

http://www.bufferbloat.net/issues/249

in addition to 243.

-- 
Dave Täht
SKYPE: davetaht
US Tel: 1-239-829-5608
http://the-edge.blogspot.com

Its the bufferbloat stupid

[Marc Manthey]

Hey dave, all

today i had the time to listen to this google tech talk again
http://www.youtube.com/watch?v=qbIozKVz73g (Bufferbloat: Dark Buffers  
in the Internet )

and i metioned the sentence from vinf cerf  in the introduction to Jim  
Gettys

"....he is working on emerging teleconferencing " whitch came to my  
interest.

I am doing videoconferencing since the early 90s and i am researching  
in the past years about large scalable secure
video conferencing a lot.

After i meet my main developer in 2008 we started to implement  
"opencu" based on the old protocol papers i´ve found.
At that time there was no efficent opensource codec and Qt wasnt  
opensourced too, so we decided to use wxwidgets and xvid / portaudio  
to implemented the basic features and made an proof of concept client.  
(The server/refector isnt re_enineered yet, only java hack available )

He got a job offer to implement such a commercial system but he failed  
after some attemts on latency and serverload.

So after i' ve tested most of all available systems like paltalk,  
ivisit, camfrog and jmeeting, webmeeting and so on , i figured  all  
solutions
whitch were mostly implemented in the 90s worked more or less worse  
and had latency about 4 seconds and were not opensource at all.

So I  decided to re_implement our social -zero-conferencing idea with  
VP8 whitch was aquired from google in 2010 and got "opensourced"
and seems highly efficent for the video part and WAB bonjour a  
zeroconfiguration service discovery framework developed
by dr.steward chesire from apple.

After i see the issues my developer had, this "unbufferbloat  project"  
came to my attention because latency and buffering seems to be the  
main issue in realtime communication plattforms.

I am looking forward to see how this problems would be solved.

regards

Marc

On Aug 25, 2011, at 4:18 AM, Dave Taht wrote:

> oops, the thing I was hoping for more eyeballs on was:
>
> http://www.bufferbloat.net/issues/249
>
> in addition to 243.
>
> -- 
> Dave Täht
> SKYPE: davetaht
> US Tel: 1-239-829-5608
> http://the-edge.blogspot.com
>

--  Les enfants teribbles - research / deployment
Marc Manthey- Vogelsangerstrasse 97
50823 Köln - Germany
skype: MarcManthey
Tel.:0049-221-29891489
Mobil:0049-1577-3329231
site: http://let.de
project : http://opencu.org
twitter: http://twitter.com/macbroadcast/

Opinions expressed may not even be mine by the time you read them, and  
certainly don't reflect those of any other entity (legal or otherwise).