# MXCuBE Developers Committee Minutes ### When: Friday, November 22nd 2024 ### Where: Trieste (Elettra Synchrotron) ### Present: P. Focia, LS\_CAT; Andrey Gruzinov, DESY; Marcus Oscarsson, Antonia Beteva, Didier Nurizzo, Yan Walesch, ESRF; Dan Costin, Martin Savko, SOLEIL; Peter Keller, Rasmus Fogh, GPhL; Annie Heroux, ex Elettra; Michael Hellmig, BESSY, Ed Daniel, Icebear; Jean Jakovic; Lais do Carmo, ESS; Alessandro Olivo, Elettra; and others ### Minutes: * Cybersecurity The meeting agrees to recommend the use of single sign-on, and reverse proxy, with a three-tier architecture as a longer-term goal (as presented in the talk of Jean-François Perrin). The meeting is in favour of having a code review done; this could cost 5-10000 Euro. It is proposed that LS-CAT might be in a position to request a review for free from the US DoE. The meeting is in favour of this, in spite of a minor worry that the acceptance criteria might differ between continents. The meeting is also generally in favour of penetration tests. It is requested that LS-CAT make a list of the most important pieces of documentation missing and notable points in general as part of the process of joining MXCuBE. The requirement that MXCuBE should run under the ID of the user who is logged in is desirable but is felt to be difficult to achieve. The meeting notes and approves that there will be cases where unusual access may be needed at short notice, e.g. for debugging purposes. * Automation Unattended data collection is a plus for MXCuBE – and a major reason for interest from LS-CAT \- but it is still not available across all facilities. It is agreed that there is a need for a clearly defined list of capabilities that a beamline must support (first and foremost automated centering). There is also a need for a standard ‘diffraction plan’ API to define how automation jobs are specified, which keywords must be supported, what is mandatory and what is optional etc. The matter will be discussed further in the MXCuBE Automation Working Group. The group will be chaired by Didier Nurizzo, and those interested (RF and MS volunteering) should join the group on the mxcubecore github. * YAML configuration The yaml configuration has been fully implemented and tested at BioMAX (with thanks to Elmir Jagudin). Testing is planned at the ESRF during the next shutdown. It is agreed to finalise the plans at the next remote developers’ meeting, and to plan a change-over to yaml configuration for the next half-yearly meeting. There is still a need to settle exactly which functions from the temporary XML/YAML implementation should be maintained and which should be purged. * Any Other Business \- It is agreed that Python 3.8 (already long out of support) should be dropped soonest. It is still in use at SOLEIL. Martin Savko is willing to upgrade to a newer version but needs to do some tests first. It is suggested to check if the version can be upgraded to a newer version than 3.9, to limit the number of changes required, and to gain access to some newer features. \- It was suggested that features like X-ray centring, reading sample resolution estimates from DOZOR, dose budget prediction, and acquisition of crystal size and shape information, could usefully be taken up for inclusion into MXCuBE. These might usefully be discussed in the first instance by the Automation Working Group. \- There are working groups also for cybersecurity, sample environment, and testing. Those interested should join the groups on github. # MXCuBE Steering Committee Minutes ### When: Friday, November 22nd 2024 ### Where: Trieste (Elettra Synchrotron) ### Participants (10): Nicola Demitri and Roberto Borghes (ELETTRA), Gerard Bricogne (GLOBAL PHASING), Daniele de Sanctis (ESRF), Judith Juanhuix (ALBA), Gleb Bourenkov (EMBL), Wenming Qin (NFPSS), Manfred Weiss (HZB), Ana Gonzalez (MAX IV), Ralf Siebrecht (ARINAX), William Shepard (SOLEIL) ### Minutes: * Next meeting will happen in Spring 2025 (second half of May?) and will be hosted in Hamburg. Gleb will discuss with Johanna regarding a joint organization and they will get back to us with a proposed date. *EDIT* (info from StC email sent on 26th November 2024): Gleb and Johanna confirm that EMBL and DESY will host the MXCuBE/ISPyB meeting and MXCuBE code camp in May 2025 in Hamburg. They propose to do it with the following structure: on May 19th to 21st; with code camp on the 19th, MXCUBE meeting 20th and ISPYB on the 21st. * The MXCuBE meeting at the end of 2025 will likely be organised by DLS/GPhL and held in the UK. It is proposed to invite some “power users” to the meeting for discussion. * Next meeting will include a thematic day with a Code Camp. A scientific thematic session will still be present but it will cover no more than half of the day. * The Steering Committee backs the decisions on YAML configuration taken by the Developers’ Committee. Status of porting MXCuBE configuration to YAML format could be presented at the next meeting. * Daniele reported that the MXCube Collaboration Agreement renewal process (signed by the parties in October 2024\) has been completed and the whole process was fast and smooth. It is valid for 5 years. * The template for the status report was reviewed and considered still valid. Everyone is reminded to stick to the template as much as possible and request a dedicated slot in the Developers and Scientific session if other developments or highlights need to be presented. * The Steering Committee is in favour of the idea of a code review and thinks the costs should be investigated. Daniele will get in contact with Jean-François Perrin (head of ESRF IT) to discuss the code audit, which should cover the latest MXCuBE code (both core and web). Daniele will also discuss with ESRF administration to find a convenient way to organize the payment * Steering Committee is favorable to LS-CAT from APS joining the collaboration. Daniele will send to them relevant documents for evaluation (i.e. Memorandum of Understanding, Collaboration agreement’ text). * If the LS-CAT team is joining, they might be required to arrange a penetration test, which could be beneficial for cybersecurity analysis. LS-CAT promises to check possibilities through DoE. It is proposed that one might have one facility do penetration tests on another, to save the cost of hiring professionals. * Automation working group chaired by Didier Nurizzo will start its activity in the coming weeks with the first meeting. Partners are requested to indicate a spokesperson if interested in participating in the discussion. It is recommended to include both scientists and developers in the Working Group. The Automation WG should report at the May MXCuBE meeting to see how far along the work has come. * More working groups coordinators will be identified at the next developers and scientists meetings. * As also recommended by the ISPyB Steering Committee, there should be a follow-up meeting to the 2019 Data Quality Metrics Workshop (DQMW) to gather input from scientists and users into software development, with contents to be made public on the DQMW web site.