There are no effective measures to protect your source.
You can, of course, run it through an obfuscation script that replaces meaningful variable names with var1 var2 var3 etc, and you can write complicated checks to see wether this is a licensed machine, but in the end, you cannot protect the source from someone who really wants to copy or modify it.
of course, you can make it hard for the average person to get your source, by not giving him access to the machine, even if it's located in his office, or putting it inside a virtual machine that he doesn't have access to, but either it's painful (virtual machine with encrypted harddisk that you have to decrypt with your passphrase whenever it is rebooted) or less secure.

Thanks janharders,

1. I have in mind putting checking procedures in certain scripts to detect any modification and then stop from working (but first the script will send me a notification).

2. I also have the idea of leaving an open door so when the scripts detects any mod and notifies me, I could shut it down via url or making the app to lock itself.

You gave me a very good idea. Instead on working with an obfuscation script I think more about encryption but not on a virtual machine basis. My first option was setting up a Linux Box and your suggestion makes me think I might have the solution on the Linux file system itself configuring the server to boot and auto login but giving no access to the files (my source code) via users and permissions. I'm researching now about the security of this approach.

I also know I can encrypt the full file system, so there goes another option. Both options offer a way to have the server running but nobody being able to access those files, it would be just as if the server was far away.

Thanks again

Ajax-ify things, keep the scripts on your server and provide SaaS.

Thanks, my app uses ajax. In this case the client wants to be able of using it locally, even with no internet (it will be a local network).

By now I'm already experimenting with user accounts and privilege administration under Linux, it seems to work, still lots to test about it.

There are 3 options of encrypting Perl code:

1. Use PAR with PAR::Filter::Obfuscate or PAR::Filter::Crypto

2. Use Filter::Crypto::CryptFile (will require some modules installed on target OS)

3. Turn into module and encrypt into Module::Crypt.

thanks chorny, checking out...