Comment 4 for bug 362142

With recent work in my trunk-dev branch, pretty much all the points in the original report are dealt with. The solutions aren't super-clean, and we still have lots of room for improvement on iptest (mainly its options handling and verbosity), but by and large the core problems reported above are fixed.

Brian, what do you think? We're never going to get off of the nose/twisted pair of tools because twisted.trial is very limited in critical areas compared to nose, yet needed for managing the reactor. So unless twisted.trial were to massively improve (extensible doctest management, full test discovery of non-unittest tests, etc), it's not a realistic option.

I'm inclined to close this bug now, leaving the remaining iptest problems to their own bug reports so we can act on them with more specificity.

OK with that?