[16145] in Athena Bugs
Re: athdir
daemon@ATHENA.MIT.EDU (Aaron M. Ucko)
Wed Aug 12 15:15:33 1998
To: Joseph Sokol-Margolis <seph@MIT.EDU>
Cc: bugs@MIT.EDU
From: amu@MIT.EDU (Aaron M. Ucko)
Date: 12 Aug 1998 15:15:31 -0400
In-Reply-To: Joseph Sokol-Margolis's message of "Wed, 12 Aug 1998 14:56:32 EDT"
Joseph Sokol-Margolis <seph@MIT.EDU> writes:
> I was actually setting more than just bindir and exec-prefix, but just
> picked one as an example.
OK...note that the locker model doesn't generally require many path
settings besides prefix (/mit/your-locker) and exec-prefix
(/mit/yourlocker/arch/$ATHENA_SYS).
> > Second, you *want* to hardcode the systype you're building new
> > software for; some aspects of athdir (for instance, its lack of output
> > if it can't find a suitable path and its ATHENA_SYS_COMPAT support)
> > are inappropriate here.
>
> Since I was actually using this in a script that was calling
> configure, wouldn't the same path get hardcoded either way? (provided
> athdir actually returns something)
Not necessarily. If you're updating an existing locker with support
for a systype in ATHENA_SYS_COMPAT but no explicit support for the
systype you're using now, you'll hardcode the incorrect systype if you
run configure before updating arch.
--
Aaron M. Ucko, KB1CJC <amu@mit.edu> (finger amu@monk.mit.edu)
