In my practical experiment it appears that it is as Gary Stanley said:
QUOTE:
"The server FPS is simply cycles / time, where time is from
gettimeofday(), with some bounds on the minimum usleep so making the
usleep actually less will crank up the server FPS (but not the
simulation HZ, so the game isn't actually faster, for Source engine games)."
END QUOTE
I`ve tested my server on kernel with no preemption and the high
preemption with cyclictest program. FPS became almost straight line when
cyclictest was showing low values (usleep stable), but how is it
connected to server quality ? I`m not sure, becouse on both kernels i
didnt notice any difference in gameplay.
Unfortunately tools like fpsmeter are missinterpretate by customers and
competiting companys, so they say often that this or that company is bad
becouse fps is not stable, which is total nonsens.
I also think that locking tickrate at some value would be good idea,
like it is in CSS.
Sorry for my english im tired :)
KN
_______________________________________________
To unsubscribe, edit your list preferences, or view the list archives, please
visit:
http://list.valvesoftware.com/mailman/listinfo/hlds_linux