schmorp at schmorp.de
Thu Dec 2 06:14:06 CET 2010
On Wed, Dec 01, 2010 at 07:57:04PM +0100, Aleksandar Lazic <al-anyevent at none.at> wrote:
> but I'am not sure if the effort to adopt the script isn't to high for
> yust a test ;-)
> I will stick to AnyEvent and don't care about POE.
OH, and do not forget that this benchmark doesn't benchmark POE performance,
but POE + AnyEvent performance.
AnyEvent has been criticised many times for its abysmally slow POE backend
(it really is slow), but not even the POE author could point out how to
get it faster without breaking it. The primary issue is lifetime: AnyEvent
doesn't want to affect the lifetime of the POE kernel, and POE can destroy
"watchers" (which are poe sessions) on its own, which is not something an
AnyEvent user expects. This is why AnyEvent uses one session per anyevent
watcher, which is POE's achilles heel.
On the other hand, session creation time doesn't affect the time it takes
to process an event, so I suspect that speeding up watcher creation will
not affect real-world performance much.
The choice of a Deliantra, the free code+content MORPG
-----==- _GNU_ http://www.deliantra.net
----==-- _ generation
---==---(_)__ __ ____ __ Marc Lehmann
--==---/ / _ \/ // /\ \/ / schmorp at schmorp.de
More information about the anyevent