Telco 20241209

Date

Monday, 9th Dec, 15:00 UTC

Connect

Agenda

Present

SB, AB, RB, HB, FS, WdN, BB, RO, BW, PC + Lukas P., Markus K., Ron H., Rubel M. (FAIRmat)

Minutes

  • domain names
    AB: We have been contacted from China that our domain.cn
    HB: we should say no to that
    AB: will ask LBL leagal team

  • PRs
    1517
    AB: PC has reviewed and approved it, so we can merge it in
    1518
    AB: HB has approved,
    RO: spelling mistakes
    AB: will resolve remaining comments, but no further requests.
    1407
    AB: the idea of the possibility for opening up enumerations is also relevant here
    HB: note that enum items should be intorduced with documentation
    AB: this has been approved in its current form, so let us do the documentation improvmenet in a separate step
    RO: we need both option of closed and open enumerations
    AB: subclassing is approved, but Lukas reported that it may not solve the problem without supporting multiple inheritance
    BW: a new vote is not necessary to revoke the request for subclassing, as it has just turned out that the implementation of the previous vote would result in a contradiction.
    AB: Let us accept it and release this requieremnt of using subclasses
    AB: NXlens_em must be approved itself before we use it in other approved definitions
    LP: It is fine to take it to a spearate PR
    AB: with these decisions, we can merge this in
    1486
    RO: documenttaion is too long and could be separated out
    AB: rendering hides the long doc and shows it collapesd
    RO: that looks good
    AB: let us remove type_oter also from here (assuming open enumerations will be supported)
    AB: after all conversations have been resolved, let us start an online vote.
    1507
    SB: FIELDNAME_xxx require the presence of a field with name FIELDNAME
    AB: clarification could be added
    AB: is this need a vote? or ot os just an implementatin of the convention?
    Telco participants decided that we do not need a vote here
    AB: let people not being on telco respond that, otherwise let merge it.
    RM: do we realy want to support nested groups?
    SB: yes, it might read confusing, but indeed the proposal supports nested groups for all those listed
    1408
    AB: fields from NXmx are pulled to the base class now, so it can be resued by others, too.
    AB: copied extra documentation (to energy from wavelength) could be replaced by simple reference.
    LP: Yes, it is OK
    PC: Are previous/next deivces needed?
    SB: Yes, we use them to store information on optical setups where *-like relations are between optical devices
    AB: let us bring this to a vote
    AB: Additionally to them, there are lots of new fields, like fluence, etc. Shall we do not put them to AppDefs
    BW: We should collect them in the base class and let AppDefs select whiuch one they need in a specific case
    HB: Ben is right. For a controlled vocabulary, it is good to collect faivoured names into base classes. Incident is needed to understand why and how we need new defs
    LP: +1
    AB: what about ultra-short pulse related charscteristics? Can we have then by using sbclassing
    LP: Yes, it is a reasonable choice
    MK: beam is not at all a straght line in EM
    AB: might not be neede to do subclassing
    WdN: are the units too restrictive
    SB: only examples are provided here
    RO: In neutron scattering, the incident polarisation, we use in % rather.
    BW: The concept defines electro-magnatic polarisation
    AB: let us go back to use the units NX_ANY fr incident and final polarisation.
    All accepted that.
    SB: Note that these are different concepts, so we may want to call them differently
    AB: In fact, we should have clear concepts
    BW: incident_polarisation_stokes is indeed defined in NXbeam
    AB: let us bring it to a vote with the comments