
6 Sep
2016
6 Sep
'16
4:34 a.m.
Occasionally a do a full validate on OSX, there are usually some tests which fail but not on other systems. I usually don't report these failures as there are always some and they are usually do to do with parts of the compiler I don't know anything about. However, on https://phabricator.haskell.org/D2485 Richard (Fung) and Ben were discussing these failures which I have also seen so it seems I would have wasted less time if I had reported these somewhere. What is the expectation for OSX? Should the test suite be entirely clean? Should I create issues when some of the tests get broken? Matt