Closed MinchinWeb closed 7 years ago
Hehe, typical. Every single new version is already failing. (sigh).
100 bonus points if you can eliminate the spurious integers from the output! ;-)
That is about how it should work!
It seems to be caught on something, as the Python 3 tests are timing out, rather than failing. As for the random numbers, they appear to be the terminal size, but I have no idea why they're printing. This calls for more investigation.
I saw those numbers (at least I think it's those same numbers) back on one of my PRs. See my comment https://github.com/CleanCut/green/pull/118#issuecomment-190324656
When I looked at the Appveyor history, I saw that "the odd "80" and "24" printed lines started between builds 273 (commit 51a709c77c655635d2a3e957f1b4f15fb773bd45) and 274 (commit 366f48ea945af5ab931c8071762cf1b5081ba118)."
I hope that helps.
Thanks @dougthor42 , that's a good lead!
OK, so the plot thickens... the tests pass when run on a single thread. I've had some weird behavior here locally (on Windows) that has been solved by dropping to a single thread.
As well, the '24's and '80' disappear, but a few [1A
(ANSI code to move the cursor up a line??) appear.
I completely forgot about this PR (sorry!).
Okay, so, only working with a single process sorta defeats a lot of the reason to use green ("fast"). Can we get Windows tests working with multiple processes?
I haven't had any luck yet :(
I'm going to go ahead and merge this. Someone really needs to figure out multiprocess stuff on Windows someday, though.
Added Python 3 support to Appveyor testing