When you are using a libfawk in your application, it is best to follow
these considerations (libfawk script users would expect these):

1. always call main() after load

The main part of the script may set up initial script states. For fbas,
it is implicitly called main() (created by the parser); for the other
languages it is the function main().

This function shall be called right after script load. The script should
not run a main loop from here (unless the application explicitly permits).
The host application may decide to unload the script on certain the
return values.

2. failing the run limit is fatal

If any call to the script exceeds the run limit, the script shall be unloaded.
The libfawk VM can not recover from run limit violation.

3. include

The host application is free to implement the include path resolution in
any way (or to simply not implement the include feature). When the
host application implements include in file system context, most
scripters would expect the file name specified for include is interpreted
as relative path to the path of the script that contains the include directive
(unless the included file name is an absolute path).

Thus best practice is to implement the relative-to-current-file mechanism.