New input or output classes may be posted to CPAN independently of ETL-Pipeline. For example, a new input class might be named ETL::Pipeline::Input::Fake. The distribution would be called ETL-Pipeline-Input-Fake.

Output classes follow the same convention. The module named ETL::Pipeline::Output::Fake in distribution ETL-Pipeline-Output-Fake.

For enhancements and bug fixes to ETL-Pipeline itself, I prefer to get pull requests rather than patches. Please do pull requests on branches.

Repository: https://github.com/rbwohlfarth/ETL-Pipeline