On Mon, Nov 18, 2002 at 04:57:03PM -0500, [EMAIL PROTECTED] wrote: > Perhaps that could be made to work if you had some way to > ensure that the user specified PREFIX was suited for use > as a rooted logical (that is not easy) then: > > ..FIRST > @ define/trans=conc PREFIX DEV:[FOO.BAR.] > > INSTALLPRIVLIB = PREFIX:[lib.perl]
Nahh. In that case I still have to leave in the special logic for when the specified PREFIX isn't suited. So it actually complicates things more. > Or if you still wanted the INSTALLPRIVLIB as before then > use something a bit like these completely untested speculative > lines: > > ..FIRST > @ define/trans=conc PREFIX DEV:[FOO.BAR.] > > PREFIX = PREFIX: > INSTALLPRIVLIB = $(PREFIX)[lib.perl] > > Or something like that (perhaps it'd be easer to > use splitdir() and catdir()?). Its easier to just leave it like it is (which uses splitdir/catdir). The real impetus was to solve this bug: http://rt.cpan.org/NoAuth/Bug.html?id=1686 I'd made the general INSTALL* logic too safe by taking into account the file concatination problem on VMS. I think I can seperate the way the INSTALL* variables are generated on VMS from everyone else by simply changing MM_Unix->prefixify() since MM_VMS has its own prefixify() now. -- Michael G. Schwern <[EMAIL PROTECTED]> http://www.pobox.com/~schwern/ Perl Quality Assurance <[EMAIL PROTECTED]> Kwalitee Is Job One grep { ref and ref !~ /^[A-Z]+$/ } kill 9, @ARGV;
