Return to BSD News archive
Newsgroups: comp.os.386bsd.development Path: sserve!newshost.anu.edu.au!munnari.oz.au!news.Hawaii.Edu!ames!elroy.jpl.nasa.gov!swrinde!cs.utexas.edu!uunet!pipex!sunic!ugle.unit.no!kari.fm.unit.no!arnej From: arnej@kari.fm.unit.no (Arne Henrik Juul) Subject: Re: modification to /sbin/ping In-Reply-To: blymn@mallee.awadi.com.au's message of 25 Sep 1993 03:17:09 GMT Message-ID: <ARNEJ.93Sep30000739@lastebil.kari.fm.unit.no> Sender: news@ugle.unit.no (NetNews Administrator) Organization: Norwegian Institute of Technology References: <1868@optigfx.optigfx.com> <BLYMN.93Sep24211709@mallee.awadi.com.au> Date: 30 Sep 93 00:07:39 Lines: 60 In article <BLYMN.93Sep24211709@mallee.awadi.com.au> blymn@mallee.awadi.com.au (Brett Lymn) writes: > From: blymn@mallee.awadi.com.au (Brett Lymn) > Newsgroups: comp.os.386bsd.development > Date: 25 Sep 1993 03:17:09 GMT > Organization: AWA Defence Industries > NNTP-Posting-Host: mallee.awadi.com.au > > >>>>> On 21 Sep 93 21:48:00 GMT, mrm@optigfx.com (Mike Murphy) said: > > > [nice idea about making ping return a 1 status on unreachable] > > Mike> > Mike> would work. This would be convenient. I entirely agree. Just do it! If you need more justification, call it "Ultrix compatibility". If the commercial vendors don't follow suit, they are not BSD (and Ultrix) compatible, but that is *their* problem. > It would work and would be convenient. This change is of the same > order that me and another guy had quite a while ago (not on a *BSD > system) about making make not require a tab before commands in rules > but rather use whitespace. All was fine until we tried to use the > make files on another system - they died badly. We relented and > un-hacked our version of make. The moral of the story is non-standard > breaks things horribly when you move out of the non-standard > environment, so if you have a script that relies on the return value > of ping and you try to use that script on a system where ping always > returns 0 you will have a broken script. However, refusing to fix broken things because they may then become useful is the dead wrong attitude. Return values is one of my pet peeves: A lot of utilities doesn't return a sensible return value on some systems. For example, "ls" always returns true (0) on some systems. This is dead wrong, and I hope it is outlawed in Posix 1003.2. I really think we should go through the utilities and fix the return status if it returns 0 on non-success. Then we can point at "existing implementation" and get it into Posix 1003.XYZ :-) > Go with the zen of life, contemplate one architecture computing and > see that history will grind you to a pulp if you get in it's way :-) > > Mike> I'm hesitant to suggest this change because the last > Mike> few people who suggested that behavior other than > Mike> stock BSD might be good were burned at the stake... > > Consider yourself mildly toasted. > > -- > Brett Lymn OK, I've got my asbestos suit. Flame away :-) - Arne H. Juul PS: I don't like the `zping' idea, either. Just fix the bug.