See also What is the major Perl successful story ?

The code on PerlMonks SOPW nodes has to be short - it exists to demonstrate bugs in a focused manner or illustrate the core ideas behind an algorithm. CUFP and Snippets are only meant for snippets. Meditations sometimes have longer code samples, usually marked RFC. But when there is a great deal of implementation code, many people only post the interface definitions or even a general statement of concept, not the actual code.

Large projects are usually found on CPAN and sometimes on SourceForge. One of the most well known is Bugzilla, but there are many, many others. For example, AWStats is used for website traffic analysis. Live Journal and Blosxum are two blogging systems written in Perl.

Many large projects probably fly under the public radar screen because their main role is behind the scenes supplying infrastructure. Moose defines an entire OOP framework for Perl. Catalyst is a web development framework. CPAN handles all of your installation and deployment. BioPerl is a rich library for bioinformatics. POE is a framework for event-handling. DBI is a framework for database integration.

And then there are the in-house systems, outside of the view of both CPAN and SourceForge. My own company has somewhere between 75K and 100K LOC of Perl code and test modules supporting various parts of our system administration, deployment and build system. These are designed to work cooperatively (we are reuse mavens) and so they function essentially as a single application with a large number of entry points. They are certainly managed that way.

Best, beth