On Mon, Dec 15, 2008 at 01:39:14PM +0100, Benjamin Wolsey wrote: > I was a bit hasty with the scuppering claim, I suppose. This would > actually support the idea that connect() should establish a genuine or > notional connection to a URL, and that URL/connection is constant until > connect() is called again, or close() (not tested). Does that sound > right?
Sounds right. It should likely support a queue of such connections and handle/advance them on heart-beat or timer. For HTTP connection the queue elements would be of AMFQueue type or something like that. When RTMP is to be added these elements might need to become of an abstract type exposing handling/advance themselves. AMFQueue::advance would then do pretty much what it currently does on 'tick' but should be augmented to also have an handle of the "notional connection" that was current at time of construction. .. or something along those lines .. --strk; GIS & Flash consultant/developer () ASCII Ribbon Campaign http://foo.keybit.net/~strk /\ Keep it simple! _______________________________________________ Gnash-dev mailing list [email protected] http://lists.gnu.org/mailman/listinfo/gnash-dev
