Telco 20260826

Date

Wednesday, 26th Aug, 15:00 UTC

Connect

Agenda

  • NIAC 2026

  • PRs

Present

SB, AB, HB, MW, GT, PC, BW, HG, RB, PM

Minutes

  • NOBUGS 2026

    • AB: same timeslots as 2 years ago (starting at 10am CET)

    • SB: schedule and contributions can be made available in indico. Let us Fabio!

  • Newcomers’ introduction

    • MW from APS: XAS

    • GT from ESS: data formats, Mc Stats

  • PRs

    • #1645: stagnant contributions

      • RB: it is to make a special directory

      • AB: why not as a subdir inside contributed definitions? Instead of stagnant, could we use archived?

      • HB: what about deprecated

      • AB: ‘deprecated’ has a special meaning in NeXus, but it could be dormant

      • SB: it needs to be checked if documentation generation still works for those definitions

      • HG: what is the workflow? when will a definition become dormant?

      • AB: could we check the lifecycle using an example

      • RB: NXsnshisto. Peter Peterson from Oak Ridge: it is not in use anymore.

      • AB: We could then mark such definitions and explain why it becomes dormant

    • #1643: half_acquisition in NXtomo

      • PC: comes from ESRF

      • MW: sound like a convenience field, because the information could be infered from other fields anyway

      • PM: can a domain specific validator check the consistency? Would it be possible to document how to do that? Or is it more documenting an intent of the data collection mode?

      • AB: Is it in a votable state?

      • BW: maybe better documentation would make it and its use clearer.

      • PM: would an ENUM be better instead of a boolean?

    • PC: we also run a PR review team to prepare for the NIAC meeting

    • #1640: NXxas and NXxas_trans (also discussed in #1352)

      • AB: sounds like a major item for NIAC 2026 because it replaces definitions.

      • SB: we should ask the community if they are really done with the developments.

      • BW: also compared to 1352 that has more modes modelled, it is not ready yet. The definition NXxas doesn’t seem to be a fit for purpose, so it should be trashed and replaced.

      • HB: it is not a good thing to remove or replace a definition because it can be problematic to use old datasets. versioning is better.

      • BW: Yes, in general, but here NXxas is not in use, so we shall not have problems with old datasets

      • AB: Versioning has been discussed, but we do not have them in NeXus

      • SB: other than the release versions, but it does not declare relationships between individual definitions in different releases.