Phil Lindsay gave a good talk about PyChecker at this month's Bay Piggies meeting.
From his work on a large project, he offered some tips on usage:
- keep PyChecker rules in a separate file, also under source control. Don't clutter the source with PyChecker directives.
- To keep track of "temporary" messages that are okay, don't create an exception for it. Rather, check in the output of PyChecker alongside the source, and diff against the prior PyChecker output. That way, any changes to the temporary issues will show up.
Phil reports that most Python projects now use PyChecker. It has become more difficult for the PyChecker team to find "un checked" code to use for testing. That's a good thing...