>> I'd be interested to see who else has this problem, and I know of at >> least one place where the problem may be coming from. It's something we >> never bothered improving because almost nobody ran into it. > > You tease. :) > > What might it be? And anything I might be able to try to solve it? >
This sounds like a problem I've never heard of before, perhaps one that has never been encountered before because of the uniqueness of your setup. The issue I had in mind involves the transport spawning multiple threads per connection to handle scheduling. Once or twice I've heard about people with lots of connections having scalability issues that we've chalked up to this. Clearly that's not the case for you, though. I wonder if it could involve the number of AIM users. I don't believe I've studied that part of the code in detail, but I have the impression that the transport maintains, in memory, a data structure per contact. As is my refrain for this list, I don't have time to look at it right now, but I there is a light at the end of the tunnel for stuff going on in the real world. Hopefully soon I'll have time to give PyAIM-t some attention. ~Chris _______________________________________________ py-transports mailing list [email protected] http://lists.modevia.com/cgi-bin/mailman/listinfo/py-transports
