-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 According to Ralf Wildenhues on 11/18/2007 8:38 AM: >> Likewise for m4. I've been using it on some systems. If there is >> another perl in the PATH, then the wrong data will be recorded. > > At best, AT_TESTED([m4 perl]) finds different versions of the programs > than the first words of $M4, $PERL point to, at worst, the testsuite > fails because either m4 or perl do not live in $PATH at all. I'm > hitting this on some test systems I have access to, where config.status > nicely saves my choice or $M4 and $PERL, but I forget to adjust $PATH > just for AT_TESTED: > > OK to just revert this?
What happens if we do AT_TESTED([$M4 $PERL]) instead? As long as we are using environment variables to override PATH searches, it seems like it should work. For $M4, it should even work when M4 was not pre-defined, since we mark M4 as a precious variable (maybe this means we should mark PERL as a precious variable for autoconf's configure.ac?). But this may also mean more work in propagating the final M4 and PERL values from config.status into the testsuite in the case that the user did not define these variables and relied on the PATH search. So, until all the work in the above paragraph is complete, feel free to revert the change as a stopgap measure. > 2007-11-18 Ralf Wildenhues <[EMAIL PROTECTED]> > > * tests/local.at: Do not test m4, perl with AT_TESTED. > - -- Don't work too hard, make some time for fun as well! Eric Blake [EMAIL PROTECTED] -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (Cygwin) Comment: Public key at home.comcast.net/~ericblake/eblake.gpg Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFHQGPk84KuGfSFAYARAgQ6AJ9TRhAyo+VsbyUEFjttBAsukhiSGgCfWsQo JnKIMZTupKzhUPwMatEMx5E= =Dn6C -----END PGP SIGNATURE-----
