Telco 20200424

Date

Friday, 24 April, 15:00 CET (+0100)

Hangout Link: https://plus.google.com/hangouts/_/j72qwlvegiojjpt3a36pfhow5ua

Agenda

• Issue #748: Mixed use of NXreflections and NXmx telco

• Issue #765: NXcanSAS “radiation” duplicates NXsource “type”

• Issue #762: [NXtomo] Need user flags

• Issue #757: How to add extra attributes and avoid conflicts with NeXus in the future?

Minutes

Present: BW, AB, FA, HG, HB, MK, PJ, PC

NIAC

• NOBUGS registration is split into separate days for the code camp

• linking not yet

• spacing for SD is fine

• need to make sure teleconferencing is adequate

Pull Request #767

• AB notes PR #767 is ready for final review.

• PJ notes that travis-ci reports a failure (reST syntax encoded within XML).

Issue #748

• NXreflections describes diffraction spots in a tabular format, having evolved from historical practice of attributing and recording diffraction spots from film-based photographs.

• it is a base class and not an appdef (currently)

• Both versions (base class and AppDef) have use cases.

• An Appdef should be written that has NXreflections as a major component so that it can be used together with NXmx AB brings up related Pull Request #767

Issue #765

• best to mark radiation as deprecated

• PJ to discuss with canSAS people

Issue #762

• user def keys should be negative

• extra (or subgroup of fields) fields (NXnotes) needed to describe user def keys

  • Need stadardised name for description

• any new positive key should be NIAC approved

AB asks MK: base classes can’t be a subentry?

• subentries must be AppDef

• entries can also be AppDef

Issue #757

• HB suggests doing what the PDB community does and have a list of registered prefixes

• starting with an underscore didn’t work out for PDB

• NIAC should grant power for approving prefix registration

• BW suggests prefixes be in uppercase to make them obvious to human readers

• PJ suggests prefixes be followed by an underscore

• PJ will post new issue with proposal for the NIAC

Any Other Business

• PC asks NXuser for a group of users

  • MK suggests using multiple instances of NXuser

• PC ask: is there a way to give numerical precision?

  • should be applicable to any general field

  • Outside of current scope - someone needs to propose a method - perhaps as an attribute

• HG asks: How has NeXus been approaching ontologies and mapping metadata betwen formats

  • NeXus predates much of the current academic interest in ontologies

  • PJ suggests talking to a group at the ALS

  • PJ asks if there is a general Ontology that NeXus can fit within

  • HG says that ontologies tend to sit next to each other as peers, rather than under an umbrella ontology

• PJ points to the index of the NeXus definitions as a starting point to mapping the NeXus ontology

  • HG will create new GitHub issue for a published NeXus ontology

  • PJ proposed using information from the manual’s index as a starting point

  • PJ noted that various items of NeXus structure (group, field, attribute) are automatically indexed

  • MK noted that position in hierarchy is important since some terms (such as name) have multiple meanings

  • ontology should include

    • NeXus class path to specify hierachy position

    • documentation content from the NXDL for each item listed

Next telco: Mid May, using a poll in the same manner as previously.