Return to BSD News archive
Newsgroups: comp.os.386bsd.development Path: sserve!newshost.anu.edu.au!munnari.oz.au!news.Hawaii.Edu!ames!agate!spool.mu.edu!bloom-beacon.mit.edu!xlink.net!scsing.switch.ch!bernina!neptune!inf.ethz.ch!weingart From: weingart@inf.ethz.ch (Tobias Weingartner) Subject: Re: modification to /sbin/ping Message-ID: <1993Sep27.115859.5299@neptune.inf.ethz.ch> Followup-To: comp.os.386bsd.development Sender: news@neptune.inf.ethz.ch (Mr News) Nntp-Posting-Host: antares.inf.ethz.ch Reply-To: weingart@inf.ethz.ch Organization: ETH - Switzerland References: <1868@optigfx.optigfx.com> <27usa4$jvd@genesis.ait.psu.edu> Date: Mon, 27 Sep 1993 11:58:59 GMT Lines: 25 In article <27usa4$jvd@genesis.ait.psu.edu>, D. Jay Newman <dn5@psu.edu> writes: > In article <BLYMN.93Sep24211709@mallee.awadi.com.au> Brett Lymn, > blymn@mallee.awadi.com.au writes: > >... 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. > > So how about if a version of ping was created called zping (or something) > which *did* return a status value? That way the script would fail in > a predicatable manner which is easily debugged (hey, get zping here!) > when put on a foreign system. > > If zping became popular, it would replace ping (however, ping would be > defined as a link to zping, for eternal compatability with older scripts). > > While this may be slightly tounge-in-cheek, the idea seems reasonable > to me. Ah, yes. But don't keep 2 versions of ping around. Hard link them, and check for a zping in argv[0]. Less disk space wasted... --Toby.