You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
As mentioned in #5324 (comment), we are passing the -s or --nocapture flag to nose when we run the tests on Travis. This causes the output of any tests to be printed immediately rather than with its failure block at the end of the run. On the one hand, this can be nice if the test run is failing catastrophically (e.g. a segfault) because we know what printed right before the process died. On the other hand, with the "running inside of gdb" functionality we have now, we will at least get a nice traceback when that happens.
I'm inclined to turn off -s, but I wanted to confirm there weren't good current reasons for keeping it on.
As mentioned in #5324 (comment), we are passing the
-s
or--nocapture
flag tonose
when we run the tests on Travis. This causes the output of any tests to be printed immediately rather than with its failure block at the end of the run. On the one hand, this can be nice if the test run is failing catastrophically (e.g. a segfault) because we know what printed right before the process died. On the other hand, with the "running inside of gdb" functionality we have now, we will at least get a nice traceback when that happens.I'm inclined to turn off
-s
, but I wanted to confirm there weren't good current reasons for keeping it on.It seems to have been added by @pelson in b8bc21f
The text was updated successfully, but these errors were encountered: