On 11/24/06, Dario De Agostini <[EMAIL PROTECTED]> wrote:
. m a r c o s a u g u s t o wrote: > why wrote a socket if we can upstream with red5? because you have to use flv if you want to use red5. If you develop your own client and server you can use what you want and the way you want it.
Hm you can send as shared object ... its data, its binary ... Its like tunneling your data....but trough red5.. I 'm thinking about multiple clients... its possible ? a multiclient server on fp9?
that's what I said in the beginning, get a screen compact using png or > jpg, and send the changed rectangles..... > but as Dario said, V7 goes further... it make the compact the using a > video codec... hmm sweet... > if FFmpeg can encode flv why red5 can't ? usually it's the client which has the coding load. switching encoding load on server has no purpose... you would need to send "raw" data to server anyway (bandwidth?).
FFmpeg has even a server for streaming..that is bugged.....maybe..hehe
> maybe... The real problem in screensharing is not the server. It's the flash client: 1) It doesn't encode with the right format 2) it can't capture desktop 3) it's slow when encoding big amount of data (at least the old v7)
yup...anyway you'll need to install something..or those java applets.... this is a fact..... but just for who wants to show .... Since you can develop your own screensharing solution with fp9 (this is
what i understand), you could solve your problem by writing your own client to grab and encode screen data. But i may be wrong... i dropped the "pursue for screensharing" long time ago :) Dario De Agostini _______________________________________________ Red5 mailing list [email protected] http://osflash.org/mailman/listinfo/red5_osflash.org
-- . m a r c o s a u g u s t o ; .eu vim para confundir e não para explicar!. . . - Chacrinha
_______________________________________________ Red5 mailing list [email protected] http://osflash.org/mailman/listinfo/red5_osflash.org
