Note: I didn't write this post to start a flame war or to discredit anyone. What i like to see is some active discussion on the topic
On most of my (linux) systems, i usually have at least two perl binaries installed:
In the past - when using the OS supplied perl - i had some major troubles. Mostly, these where incompatibilities after the distributor rolled out some updates or when i upgraded some modules to newer (test) versions. Also, now i use ActiveState Perl for much of my software development.
The trouble is: While the distribution-supplied scripts correctly call /usr/bin/perl, most of the code that i download from the interwebs does so too - and start breaking because of "missing" modules
The thing that worked for me best is to change all Perl scripts on my system to use
#!/usr/bin/env perl
as the hashbang line. Whenever the systems automatically start a Perl script, it uses the dristribution-maintained perl, whenever i start a script with my user account, it uses the perl installed in my home directory.
In my opinion, in a time where there are more and more different plattforms, assuming perl is at a fixed path and that perl binary is the correct one to use isn't very helpfull. It's similar to the still widely spread believe that /bin/sh is bash (or csh or whatever)...
In reply to Calling the correct perl binary by cavac
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |