make test fails
Closed this issue · 4 comments
I also found a possible fix for make test. If to replace init_test with init in TESTS = ... everything should be ok. In this issue we should check what is the difference between sql/init.sql and sql/init_test.sql. May be, sql/init.sql is enough?
I have a fix for the USE_HEALPIX=0 case, but I can't compile the code with USE_HEALPIX=1 to get the expected output in that scenario. If someone could compile PR #26 with HEALPix support and upload the regression.diffs file for make test, I could include include the fix for make test in PR #26, but maybe it should be a separate PR?
Oh, I just realized I could leverage the Travis CI to do this! I'll try doing that.
I also found a possible fix for make test. If to replace init_test with init in TESTS = ... everything should be ok. In this issue we should check what is the difference between sql/init.sql and sql/init_test.sql. May be, sql/init.sql is enough?
Oh, interesting! That would simplify things, certainly. I'll take a look.