https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=277826
--- Comment #40 from Max Brazhnikov <[email protected]> --- (In reply to Charlie Li from comment #38) The report in https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=277826#c28 is not relevant. I repeat myself, you are not supposed to replace Numpy 1 with 2 on the fly. Stick to Numpy 1 if you have built other ports against it. > While the two are not firmly coupled, in reality this needs to run with numpy > 2 Upstream disagrees with you. Have you tried it yourself? > Splitting docs creates more unnecessary maintenance work. It's the opposite for me :). I had this change in mind before the update. The main motivation is to avoid multiple copies of the same docs when you install more then one flavor. > For those adventurous on testing with numpy 1 in the meantime, adjust the > relevant dependency lines accordingly, run the test suite and report the > results. ============================================= short test summary info ============================================= FAILED scipy/optimize/tests/test_zeros.py::TestNewton::test_halley_collections - AssertionError: FAILED scipy/signal/tests/test_filter_design.py::TestBessel::test_degenerate[numpy] - AssertionError: FAILED scipy/stats/tests/test_continuous.py::TestDistributions::test_logexp_safe[logcdf] - assert -0.36894641528865635 != -0.36894641528865635 ================================== 3 failed, 69584 passed, 7576 skipped, 11661 deselected, 208 xfailed, 20 xpassed in 867.47s (0:14:27) ======= -- You are receiving this mail because: You are the assignee for the bug. You are on the CC list for the bug.
