On Fri, Apr 8, 2011 at 3:31 PM, Steven McCoy <steven.mc...@miru.hk> wrote:
Thanks for the code review! > I think you've been beguiled by the Windows API, unsurprisingly in reality > it is pretty terrible. :) of course. > If you call Sleep() with zero milliseconds you end > up with a yield on older versions of Windows. Also, you are going to get > only 16ms resolution and you must use timeBeginPeriod() to actually get to > 1ms. Send me a patch? > On Unix as well as nanosleep() you have access to usleep() for microseconds. > http://pubs.opengroup.org/onlinepubs/007908799/xsh/usleep.html usleep is deprecated and removed from POSIX.2008. > This doesn't do what you think it does, use GetSystemTimeAsFileTime() > instead. Hmm, I vaguely remember this from a previous life. Send me a patch? I've no Win32 box booted right now. > I'm surprised you don't have a zclock_delay() API, but I guess depends on > demand. What would this do? zclock_sleep() already exists. -Pieter _______________________________________________ zeromq-dev mailing list zeromq-dev@lists.zeromq.org http://lists.zeromq.org/mailman/listinfo/zeromq-dev
