From owner-freebsd-bugs  Thu Jul 24 05:35:23 1997
Return-Path: <owner-freebsd-bugs>
Received: (from root@localhost)
          by hub.freebsd.org (8.8.5/8.8.5) id FAA25554
          for bugs-outgoing; Thu, 24 Jul 1997 05:35:23 -0700 (PDT)
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19])
          by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id FAA25546
          for <freebsd-bugs@hub.freebsd.org>; Thu, 24 Jul 1997 05:35:20 -0700 (PDT)
Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.5/8.6.9) id WAA20266; Thu, 24 Jul 1997 22:25:52 +1000
Date: Thu, 24 Jul 1997 22:25:52 +1000
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199707241225.WAA20266@godzilla.zeta.org.au>
To: jkh@time.cdrom.com, TLiddelow@cybec.com.au
Subject: Re: bin/4154: wish /bin/sleep handled fractions of a second.
Cc: freebsd-bugs@hub.freebsd.org, thorpej@nas.nasa.gov
Sender: owner-freebsd-bugs@FreeBSD.ORG
X-Loop: FreeBSD.org
Precedence: bulk

>In this particular case, if you have a script which says something like:
>
>foo
>sleep 0.8
>bar
>sleep 0.9
>baz
>
>And you bring it to a non-BSD system, it will not sleep _at all_ since
>the other system sees "sleep 0", and that could be bad depending on
>what bar and baz do.

It should of course do something like:

sleep: invalid time `0.8' (not a nonnegative decimal integer as specified
       by POSIX.2 4.57.4)
sleep: invalid time `0.9' ...

BSD's sleep has the usual sloppy numeric arg checking using atoi(), so it
won't do anything like this.

>This is exactly the kind of interoperability
>problem that POSIX was intended to try and solve.  Let's not fight it.

I agree.

Bruce