[2128] in SIPB bug reports
where
daemon@ATHENA.MIT.EDU (Ken Raeburn)
Wed Oct 2 18:19:59 1991
Date: Wed, 2 Oct 91 18:15:51 -0400
From: Ken Raeburn <Raeburn@MIT.EDU>
To: ckclark@MIT.EDU
Cc: bug-sipb@Athena.MIT.EDU, marc@Athena.MIT.EDU
In-Reply-To: <9110022056.AA08789@hal-2000.MIT.EDU> (message from Calvin Clark on Wed, 2 Oct 91 16:56:18 -0400)
Date: Wed, 2 Oct 91 16:56:18 -0400
From: Calvin Clark <ckclark@Athena.MIT.EDU>
But you're right, there is a builtin `where' in the current Athena tcsh,
and its use in scripts does not produce the same ioctl error that the
built-in which does, so part of my argument for keeping is blown away,
unless you want to use it in a Bourne shell script. The proper defense
for the program would have to come from someone who's actually used it a
lot.
Well, that'd be me. I used it quite a bit before I switched to tcsh,
and will continue to want to use it on machines that don't have tcsh
(or some other shell supporting exactly that functionality) as
/bin/csh (unless or until Athena figures out some way of making a
user's shell choice architecture-dependent).
I also don't like having a new shell started up to source .cshrc, and
don't want to have to remember to use different names under different
shells for such an important feature. And using it in a Bourne shell
script (e.g., for a command that will get run in lots of places
through the script) seems like a good reason also.
(Does tcsh-6 fork and re-read .cshrc to execute its "which" builtin,
or use the current path?)
The extremely trivial shell script "ds" in /usr/sipb also uses this
program; I use this on occasion, but it would seem like a more likely
candidate for elimination.
