pyiron_base broken!?

[niklassiemer]

I tried to increase the version number of pyiron_base in #221 and this breaks our tests! (see below)

@pmrv I fear this is related to the changes in the copy_to we made in base? Since the tests for the new pyiron_base release are failing and I am in holidays from tomorrow on, could you please take over this release cycle? @muh-hassani please follow the discussion before updating the packages.

Traceback (most recent call last):
full_job_sample full_job_sample
  File "/home/runner/work/pyiron_atomistics/pyiron_atomistics/tests/vasp/test_vasp.py", line 345, in test_run_complete
job was stored with the job ID  1
    self.assertTrue(h_out.list_nodes() == [])
full_job_minor_glitch full_job_minor_glitch
AssertionError: False is not true

[liamhuber]

Did the new hierarchy interface get included in this release? Given that a .list_nodes call is failing, I am suspicious.

Also, I think there is assertListEqual, which might give us a more useful error message on failure.

[niklassiemer]

No, we do not have the new .list_nodes interface, yet. In this case actually also just an self.assertEqual would help to see what is actually given back by the list_nodes.

[niklassiemer]

I made a PR at our conda-forge feedstock to strictly fix to the older pyiron_base version to prevent breaking the environment for all pyiron_atomistics users.

[pmrv]

Seems like the test is checking that the output was not copied, but it did. So it's almost surely a bug from me in copy_to.

[pmrv]

Fixed in here.

[niklassiemer]

Thanks a lot! Let's do a new release of base and update our dependencies 🙂

[pmrv]

conda update is running right now, I'll bump the version to >= 0.2.15 tomorrow morning then.

[niklassiemer]

Perfect! Once base is in, a conda install pyiron(_atomistics) will give a working base! Thanks a lot for taking care of this!

[pmrv]

New release is tagged.