From owner-cvs-lib  Fri Aug 16 04:49:00 1996
Return-Path: owner-cvs-lib
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id EAA26996
          for cvs-lib-outgoing; Fri, 16 Aug 1996 04:49:00 -0700 (PDT)
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19])
          by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id EAA26873;
          Fri, 16 Aug 1996 04:45:17 -0700 (PDT)
Received: (from bde@localhost) by godzilla.zeta.org.au (8.6.12/8.6.9) id VAA19635; Fri, 16 Aug 1996 21:37:04 +1000
Date: Fri, 16 Aug 1996 21:37:04 +1000
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199608161137.VAA19635@godzilla.zeta.org.au>
To: j@uriah.heep.sax.de, jkh@time.cdrom.com
Subject: Re: cvs commit: src/lib/libncurses lib_newterm.c
Cc: CVS-committers@freefall.freebsd.org, cvs-all@freefall.freebsd.org,
        cvs-lib@freefall.freebsd.org, jkh@freefall.freebsd.org,
        joerg_wunsch@uriah.heep.sax.de
Sender: owner-cvs-lib@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

>> > Why is this?  sigaction() is Posix, signal() ain't.
>> 
>> sigaction had unwonted side-effects.

Actually, signal() has wanted side effects.

>I don't trust you. ;-)  signal()'s implementation is mainly based on
                                                      ^^^^^^ completely
>sigaction().  So if the latter has unwanted side-effects for you that
>signal doesn't have, it's probably a programmer's error.

The only difference (under BSD) is that signal() sets SA_RESTART for
most signals.  The set of such signals may be changed using
siginterrupt().

Under SysVish systems, signal() also sets SA_RESETHAND and SA_NODEFER
to give traditional unreliable signals.  signal() should only be used
if you want traditional system-dependent behaviour, i.e., almost never.

Bruce