On Sun, Nov 11, 2007 at 11:59:26PM -0800, Niels Provos <[EMAIL PROTECTED]> wrote: > - performance improvements due to using a heap instead of red-black > trees for timeouts
I did the same benchmark as against 1.3e (see http://libev.schmorp.de/bench.html). The two graphs are available here: http://libev.schmorp.de/1.4.0-beta.t0.png no timers http://libev.schmorp.de/1.4.0-beta.t1.png timers (its a bit more spiky as usual as I used less benchmark runs and also did some surfing while waiting for the resuls :) as you can see, the heap indeed pays off, the growth behaviour is now the same as with libev (unlike my earlier test just after heap integration, which turned out to be quite a bit slower). at lower number of timers, libevent has virtually the same behaviour as before, at higher numbers it is quite a bit better now! -- The choice of a Deliantra, the free code+content MORPG -----==- _GNU_ http://www.deliantra.net ----==-- _ generation ---==---(_)__ __ ____ __ Marc Lehmann --==---/ / _ \/ // /\ \/ / [EMAIL PROTECTED] -=====/_/_//_/\_,_/ /_/\_\ _______________________________________________ Libevent-users mailing list Libevent-users@monkey.org http://monkeymail.org/mailman/listinfo/libevent-users
