[comp.unix.wizards] Csh brain damaged, but then you already knew that.

dmcanzi@watdcsu.waterloo.edu (David Canzi) (09/16/89)

This happened on a vax running 4.3BSD or something like it:  

watdcsu % history | tail -5
tail: : open of temporary file "5.a" failed -- Permission denied.

It also happened on a Sun running SunOS4.0.3:

shine22% history | tail -5
tail: : open of temporary file "7.a" failed -- Read-only file system.

The problem seems to be that csh tries to create a temporary file in
the current directory, and in the above two cases I didn't have the
necessary permissions.

Feh.

-- 
David Canzi

dmcanzi@watdcsu.waterloo.edu (David Canzi) (09/16/89)

Sorry.  False alarm.

The problem turned out to be the fault of the local version of the
"tail" command.

My final comment, however, still applies:

>Feh.

-- 
David Canzi

guy@auspex.auspex.com (Guy Harris) (09/17/89)

 >This happened on a vax running 4.3BSD or something like it:  
 >
 >watdcsu % history | tail -5
 >tail: : open of temporary file "5.a" failed -- Permission denied.
 >
 >It also happened on a Sun running SunOS4.0.3:
 >
 >shine22% history | tail -5
 >tail: : open of temporary file "7.a" failed -- Read-only file system.
 >
 >The problem seems to be that csh tries to create a temporary file in
 >the current directory, and in the above two cases I didn't have the
 >necessary permissions.

Wanna bet?  I just tried running "history | tail -5" under "csh" under
SunOS 4.0.1 in a directory to which I didn't have write permission, and
it worked just fine.

What's more, I couldn't find any error message string like that in the
source to the 4.3BSD "csh" *NOR* in the source to the 4.3BSD "tail", and
"strings" didn't find it in the SunOS 4.0.1 binaries, either.

Maybe you've got some mutant version of "csh" or "tail" there, and
*that* version is the putatively "brain-damaged" oone?