Can one do similar to --nocapture with discover_road_runner.runners.DiscoverRoadRunner

[jorourke]

Is it possible to see console output while running tests?

[pzrq]

@jorourke Simple answer is probably no.

Longer answer is I haven't really hacked on this for a couple of years. I honestly don't recall if --verbosity=2 or greater (which sounds like it may be closer to nose's --nocapture?) in Django's test runner based on DiscoverRunner on which this DiscoverRoadRunner was based worked reliably because of the inherent interleaving of output streams from the different Python processes.

When last I made time, Django 1.9 introduced a --parallel option:
https://docs.djangoproject.com/en/1.10/topics/testing/overview/#running-tests-in-parallel

At that point if I continued to need this as a workaround for not paying for TravisCI (yes it's as dumb as it sounds, but it's sometimes fun to make suboptimal choices because you can learn a lot) I'd have rewritten it building on top of Django's built in runner.

I would like some of the ideas, especially the app-based traffic light feedback, to be built into Django or a reasonable drop in replacement rather than this pile of hacks, but I acknowledge that at the end of the day, it's the kind of thing that's really only useful for driving massive change over a heavily unit-tested monolithic codebase, such as a complete Python 3 conversion. However that seems far from a common use case because moving incrementally in small steps is almost always better for a project over time, and if you move incrementally in small steps you simply don't break 500 tests at a time and need to get a feel for what in your project is really broken and where the easy 450, then 45, then the last diabolical 5 wins are.

[jorourke]

@pzrq , did you work at mathspace a while back? Small world if so. Started there a while back. Ok, thanks for the response. Not a top priority.