True. For a standalone script, it's always nice to see some globally used parameters at the top for easy customization/configuration.

I do the follwing in such cases...

If the variables are "configuration" type of variables, I would use constants, not plain scalars. I feel much better in knowing that most ( if not all ) of the globally accessible variables would be visually different from other local variables:

   #!perl

   use constant LOG_DIRECTORY => '/foo/bar/baz';
   use constant SOME_URL => 'http://baboon.com';

   main();

   sub main {
     ....
     some_random_function( LOG_DIRECTORY, $local_variable );
   }
[download]

using constant is a very good thing to do. However, if you don't know what constant is doing, you can be very surpsised.

use constant LOG_DIRECTORY => '/foo/bar/baz';

# Equivalent to

sub LOG_DIRECTORY () { '/foo/bar/baz' }
[download]

The most obvious problem is interpolation in double-quoted strings.

Similarly, you can't use LOG_DIRECTORY as the key to a hash, because it's not a C-style preprocessor declaration, but a function declaration.

Just, be careful and know what is happening under the hood.

------
We are the carpenters and bricklayers of the Information Age.

Don't go borrowing trouble. For programmers, this means Worry only about what you need to implement.

Doh! Nevermind....

I can see the interpolation problem, but whats this about not using a constant as a hash key?

#!/usr/bin/perl -w
use strict;

use constant LOG_DIRECTORY => '/foo/bar/baz';

my %hash;
$hash{LOG_DIRECTORY} = 'happy';

print "log directory ${\LOG_DIRECTORY} is $hash{LOG_DIRECTORY}\n";

__END__
produces:

log directory /foo/bar/baz is happy
[download]

-Blake

Try printing out the keys of %hash. You just assigned the value "happy" to the key "LOG_DIRECTORY". To use the constant LOG_DIRECTORY as the key you need to explicitly use it as a function.