Unpin lalsuite version #5040
Unpin lalsuite version #5040
Conversation
|
Don't you also need to change something here? https://github.com/gwastro/pycbc/blob/master/.github/workflows/basic-tests.yml#L31-L37 |
|
Huh, HDF5 errors and a segfault, interesting… |
I think they're the same error, and I think it originates with me not understanding git lfs ... Grumble, grumble, grumble ... Moving to a direct curl instead, I understand curl. |
|
But the segfault is indicative of a bigger problem, potentially an unchecked error condition in lalsimulation. |
That's probably right .... But I'm not sure I want to go down that rabbit hole. Jolien's HDF code did fail, so I suspect that SEOB carried on with a |
The various data files needed for lalsuite are now available publically again (lalsuite will inform the user of this if the user does not have the files in their path).
Therefore I unpin lalsuite. I've dumped the new SEOBNR file into our existing CVMFS directory so the CI will need no more changes.
To confirm SEOBNRv4_ROM_v3 is identical to SEOBNRv4_ROM_v2 (it just contains a certain bit of metadata that is irrelevant for us).
This would be needed for #5019 (and a lalsuite release that uses igwn-ligolw, if that doesn't already exist).