Open seisman opened 5 months ago
Yeah, seems like these same tests were failing in our GMT Dev Tests since we bumped to macos-14 as mentioned at https://github.com/GenericMappingTools/pygmt/pull/3024#pullrequestreview-1858990667, see https://github.com/GenericMappingTools/pygmt/actions/runs/7764289384/job/21177511396.
In GenericMappingTools/gmt#8188,
if (tx_a < ta_stop && tx_b < tb_stop)
was changed toif (tx_a < ta_stop && tx_b <= tb_stop)
. I guess the floating-point comparison may be tricky and depend on the C compiler.
Should we xfail this then on macOS, since the floating point comparison is consistent on Linux and Windows? Or pin to e.g. macos-13
for now (though this only delays the issue).
Should we xfail this then on macOS,
i think yes
https://github.blog/changelog/2024-04-01-macos-14-sonoma-is-generally-available-and-the-latest-macos-runner-image/
macos-latest
has migrated from macOS-12 to macOS-14 in a few hours ago. All tests are good exceptx2sys_cross
tests.The failures are:
For reference, the data shapes were updated in PR https://github.com/GenericMappingTools/pygmt/pull/2986 after the upstream fix https://github.com/GenericMappingTools/gmt/pull/8188.
In https://github.com/GenericMappingTools/gmt/pull/8188,
if (tx_a < ta_stop && tx_b < tb_stop)
was changed toif (tx_a < ta_stop && tx_b <= tb_stop)
. I guess the floating-point comparison may be tricky and depend on the C compiler.