Message125628
On xp, changing from -m test to -m test.regrtest removed the extra craziness during and after the test run that I reported on pydev.
I think making at least a tempory fix to whatever -m test runs should be a release blocker so only individual tests and not the test process are broken. If necessary, just refuse to run under windows and say to use test.regrtest as before. (I believe I did use -m test.regrtest on 3.1 and 2.7 as -m test is not implemented for those.) |
|
Date |
User |
Action |
Args |
2011-01-07 08:03:40 | terry.reedy | set | recipients:
+ terry.reedy, brett.cannon, ncoghlan, pitrou, vstinner, jnoller, michael.foord, brian.curtin, asksol |
2011-01-07 08:03:40 | terry.reedy | set | messageid: <1294387420.34.0.83609251304.issue10845@psf.upfronthosting.co.za> |
2011-01-07 08:03:36 | terry.reedy | link | issue10845 messages |
2011-01-07 08:03:36 | terry.reedy | create | |
|