Author pitrou
Recipients belopolsky, neologix, pitrou, serhiy.storchaka, skrah, vstinner
Date 2014-02-11.23:19:18
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <1392160756.2292.7.camel@fsol>
In-reply-to <1392160653.45.0.507810788686.issue20320@psf.upfronthosting.co.za>
Content
> I consider that sigtimedwait() is similar to select and kqueue: it's a
> function to wait for an event with a timeout. IMO it should use the
> same rounding method. Don't you think so?
> 
> I don't really care how sigtimedwait() rounds its timeout, I just
> tried to be consistent.

We have lots of "waiting for an event" with a timeout function (e.g.
Lock.acquire()), so if we want to be consistent we have to change all of
them.
But originally this issue only "affects" event loops, so presumably it'd
be reasonable to stick to select()-alikes.
History
Date User Action Args
2014-02-11 23:19:18pitrousetrecipients: + pitrou, belopolsky, vstinner, skrah, neologix, serhiy.storchaka
2014-02-11 23:19:18pitroulinkissue20320 messages
2014-02-11 23:19:18pitroucreate