Aaron S. Brewster

Chiming in, that does look like it would break the stills indexer in multi-lattice mode :)

@ndevenish ok to merge? The xfel handler ok? Thanks and no hurry :)

Some folks are on leave but I think this is fine. Ima hit the auto-merge button. Thanks @dermen.

Chatting with @bkpoon, this comes from https://github.com/python/cpython/issues/80266 If understood right, a cctbx change needs to be made which will propagate to fixing these problems. Something something library compilation order :)

Hi, the phil params are shown in the log file for `dials.stills_process`, named `dials.process.log`. That happens [here](https://github.com/dials/dials/blob/main/src/dials/command_line/stills_process.py#L578-L581). Is this sufficient for the phil params? The other bits you've added are...

(gonna let the tests cook)

Hi, mpi is used in lots of xfel_regression tests. https://github.com/dials/dials/blob/d7062646c8c999ba4bf68eaa14e2e37f901e0f8e/.azure-pipelines/unix-build.yml#L114 Could openmpi be dropped from this line? mpi4py will bring in either openmpi or mpich, depending on which satisfies the...

Ok, looking at this today. I found that at the top of the log for the original version it says this: ``` Spots on d-rings: 12 Total sf spots: 14...

Ok, I've found the source of the 10 vs. 12 spots on rings, the wavelength. Use `use_beam_reference=False` for dials.import: ``` dials.import ../ten_cbfs/cxily6520_r0164_02081.cbf reference_geometry=split_0.expt use_beam_reference=False ``` The next problem seems to...

Got it! indexing.refinement_protocol.mode=None was part of the answer, but the full command that worked is ``` dials.index imported.expt strong.refl method=small_cell unit_cell=6.0226,7.2623,29.5203,90,90,90 space_group=C222 indexer=sequences indexing.refinement_protocol.mode=None refinement.parameterisation.crystal.fix=cell min_nref_per_parameter=1 reflections.outlier.algorithm=null ``` This requires...