[12292] in Athena Bugs
sun4 7.7C: /bin/true
daemon@ATHENA.MIT.EDU (Craig Fields)
Thu Jul 28 01:16:12 1994
To: bugs@MIT.EDU
Date: Thu, 28 Jul 1994 01:15:42 EDT
From: Craig Fields <cfields@MIT.EDU>
System name: perilous
Type and version: SPARC/Classic 7.7C
Display type: cgthree
What were you trying to do?
Make fun of the well-known absurdities of /bin/true.
What's wrong:
I was making fun of the revision number:
#ident "@(#)true.sh 1.6 93/01/11 SMI" /* SVr4.0 1.4 */
I wondered if we had an earlier version in the last
release, so I diffed them and found:
diff /afs/athena/system/sun4c_51/srvd.76/bin/true \
/afs/athena/system/sun4c_53/srvd.77/bin/true
1c1
< #!/sbin/sh
---
> #!/usr/bin/sh
9c9
< #ident "@(#)true.sh 1.5 92/07/14 SMI" /* SVr4.0 1.4 */
---
> #ident "@(#)true.sh 1.6 93/01/11 SMI" /* SVr4.0 1.4 */
Hey! We have a change! What's the difference? (Some people
know what this means without looking, but let's be clear.)
file /sbin/jsh
/sbin/jsh: ELF 32-bit MSB executable SPARC Version 1, statically linked, stripped
file /usr/bin/jsh
/usr/bin/jsh: ELF 32-bit MSB executable SPARC Version 1, dynamically linked, stripped
That is, /bin/true requires your dynamic libraries to be in
working order. I suspect this is a bug, and wonder if Sun really
considered this change carefully. They did not provide a different
version of true for /sbin. Seems risky.
What should have happened:
Maybe nothing and this is correct.
Or possibly, /bin/true should be static. Or, a static true
should be included in /sbin.
Arguably, this is not a bug for /bin/false.
Please describe any relevant documentation references:
true(1): "true does nothing, successfully."
