More HP-inspired patches

The Hermit Hacker scrappy at hub.org
Wed Sep 15 20:11:39 UTC 1999


On Wed, 15 Sep 1999, Per Hedeland wrote:

> Heath Kehoe wrote:
> >The main reason I suggest using sigaction() if it is available, is
> >that its behavior is more consistent across different OSs than signal()
> >is (since it's a POSIX-defined function).
> 
> FWIW, I agree 100% - no Unix application should be using anything other
> than sigaction() (which may well be wrapped for calling convenience and
> other reasons, of course) in this day and age, unless absolutely forced
> to due to sigaction() being unavailable - which isn't the case for any
> modern Unix. The fact that INN still uses signal() can just barely be
> excused with the old "historical reasons".

Okay, this I didn't know...and your arguments for sigaction() are
good...my original objection revolved around the comment that the
HAVE_SIGACTION was being used to replace an unreliable signal() in HP/UX,
but was forcing everyone else to use it that has it...

I retract my earlier objection...:)

Marc G. Fournier                   ICQ#7615664               IRC Nick: Scrappy
Systems Administrator @ hub.org 
primary: scrappy at hub.org           secondary: scrappy@{freebsd|postgresql}.org 



More information about the inn-workers mailing list