On Wed, Oct 20, 1999 at 03:24:01PM -0700, Randal L. Schwartz wrote: >>>>>> "Ed" == Ed Peschko <Ed_Peschko@csgsystems.com> writes: > >Ed> How difficult would it be to add the path of the current working directory >Ed> of the script itself into INC? I'm finding myself constantly doing the following >Ed> hack: > >It cannot be done with certainty. And yes, Tom's post shows that >there already is a way to do it with some uncertainty. > >I would be opposed to putting it in because it's expensive to do, >and not used for 99% of my programs. :) *sigh*. Ok, your point is well taken, but sometimes I *really* wish that there was something like a PERL_HEADER variable in which you could put stuff like this, and which would be included at the beginning of each file. Yes, there would be a performance hit, and yes it has the potential to make things non-portable, but its really beginning to SUCK to have to add six+ lines at the beginning of every file just to get functionality that is so common but also site-specific. Stuff like: use strict; use DBI; use FindBin; use lib $FindBin::Bin; use IO; use Data::Dumper; use Rebol; use lib "put site specific libraries here"; And sometimes, I wish that there was a method for *compiling* these commonly used modules into the interpreter itself, as part of Configure. I swear that having this feature would get rid of 60% of the 'should this functionality be in the core' arguments that are on p5p. Ed