From owner-freebsd-bugs  Fri Apr 21 02:47:18 1995
Return-Path: bugs-owner
Received: (from majordom@localhost)
          by freefall.cdrom.com (8.6.10/8.6.6) id CAA04922
          for bugs-outgoing; Fri, 21 Apr 1995 02:47:18 -0700
Received: from gwdu17.gwdg.de (gwdu17.gwdg.de [134.76.10.98])
          by freefall.cdrom.com (8.6.10/8.6.6) with SMTP id CAA04916
          for <bugs@freebsd.org>; Fri, 21 Apr 1995 02:47:14 -0700
Received: from namu01.gwdg.de (actually namu02.gwdg.de) by gwdu17.gwdg.de 
          with SMTP (PP); Fri, 21 Apr 1995 11:51:27 +0200
Received: by namu01.gwdg.de (5.57/Ultrix3.0-C) id AA18735;
          Fri, 21 Apr 95 11:45:09 +0200
Date: Fri, 21 Apr 95 11:45:09 +0200
From: rw@namu01.gwdg.de (Rainer Wittmann UMS)
Message-Id: <9504210945.AA18735@namu01.gwdg.de>
To: bde@zeta.org.au, bugs@FreeBSD.org, rw@namu01.gwdg.de
Subject: Re: signal handling bug
Sender: bugs-owner@FreeBSD.org
Precedence: bulk

It's not a bug. It's a feature!
Especially, if it is combined with siginterrupt, it is more flexible
and much more natural than SYSV `signal', while it still can be used
without man pages, which cannot be done with sigaction. The rather
complicated and powerful sigaction seems to be the only portable
way. I did know about it before, but I thought that it is too complicated
to be portable and stayed away.
Many thanks to all for kindly telling me the facts and not flaming me. 
I did make test before on Linux, AIX and DEC OSF/1.