In addition to
gjb's comments, if you are worried about requring too much 'stuff' that may not be needed during any one run, you could eval your commands and build your require/use list accordingly:
if($command = 'getuser') { eval { use MyCode::Users; }; getuser(); }
That way if you break up your code into multiple scripts and require them (as per
gjb's comments), you are still only loading that which you need, dynamically. If you have a lot of subs and a lot of 'commands', this can require some precise forethought, but a little design could save you lots of resources in the long run.
