OPTIMIZE_TIMEOUTS avoids extra alarm() system calls to manage network
timeouts... and alarm() shouldn't be used in 2.0 anyhow since all network
timeouts should be implemented through poll(). so OPTIMIZE_TIMEOUTS
should not be necessary in 2.0.
-dean
On Tue, 27 Mar 2001, Jeff Trawick wrote:
> Does anybody understand 1.3 OPTIMIZE_TIMEOUTS enough to state with
> confidence whether or not it makes sense with 2.0?
>
> I think that at best it could make sense only with prefork (sending
> SIGALRM to wake up child threads is definitely not possible).
>
> I don't understand the 1.3 code enough to tell what is being
> optimized.
>
> Thanks...
>
> --
> Jeff Trawick | [EMAIL PROTECTED] | PGP public key at web site:
> http://www.geocities.com/SiliconValley/Park/9289/
> Born in Roswell... married an alien...
>
