DIALS core meeting 2021-02-11

Previous Actions

• MG+ND: write a proposal for ‘overall architecture discussion’ and ‘new installer’
  • MG+ND: Come up with a proposal to move away from all code being in header files and consolidate into a single library
  • Will discuss new installer on the 23rd at 10:00
• MG: organise a typing intro lecture, likely to happen in Q2.
• cbflib conda-forge package
  • ND to prepare an agenda for the HDRMX meeting in March
  • ASB to organise a side-meeting next to the HDRMX workshop
  • ASB to explore what should happen to cbflib maintenance in the future
• ND to set up a cbflib_adaptbx pull request for MAR detector fixes. Verify with labelit_regression.
• ASB to review dxtbx#289
• ASB to run tests on dxtbx#288 as per comment
  • MG to send an email to dials-support on how to avoid UserWarning
• PR dials#1567
  • ASB to add xfel test instructions to the PR
  • Then GW to run tests and merge if passing
• dx2 proposal
  • ASB to reply on list
  • GW to merge all the different discussion strains together
  • MG to put in a PR to dx2 turning the readme to a proposal

Agenda

Construct Experiments directly rather than via DataBlocks #288

(already has a merge conflict)

This was discussed in the open action points above.

dials.index: create new experiments

dials.index should (or should not) create new experiments with crystals in rather than modifying in place existing experiments as proposed in this issue (#1029)

• Q: This would double the size of serial experiments?
• A: It wouldn’t double the size, but increase the size slightly, and there are more pointers.
• The impact of this could be tested before work on this is done.

ASB ran this test during the meeting and added results to the ticket. Results look favourable.

• Q: Is there any further objection to this if the performance impact isn’t a blocked?
• A: No obvious objections. This could lead to downstream work needed to clean up afterwards.

Change in behaviour would only affect dials.index command line program, not the API call (thus not stills processing).

If dials.index gets a single experiment and can’t find a lattice - should it write out the input file as is – ie. not add a lattice, not reassign experiments, #1080? Sentiment is that it should definitely still terminate with a non-zero exitcode for scripting purposes. Writing out the input file means users may pass the file on to dials.refine and get hit with an even less informative error message.

Renaming master → main

xia2 has now renamed its main branch. Plan is to do the same with DIALS and dxtbx in <2 weeks time.

Next meeting

Thursday, February 25th, 4pm UK time, 8am PDT.