"Craig A. Berry" <[EMAIL PROTECTED]> writes:
>
> Isn't the problem that if MMS TEST defines or redefines PERL_ROOT and
> you then proceed (as would ordinarily be the logical next step) to do
> an MMS INSTALL you will be installing the build tree on top of itself
> rather than in wherever PERL_ROOT originally pointed?

You're safe with MM[SK], since each invocation spawns off a subprocess
in which the commands get executed, so process logicals won't propagate
from one run to the next.  You may get bitten my some versions of Make
which build a DCL procedure and then execute it in the current process.
(Dunno whether anyone's using them; it's been a _long_ time since we
supported make for VMS perl.  I hope David Denholm's still well out
there somewhere.)

Regards,
Charles Bailey  [EMAIL PROTECTED]

Reply via email to