Draft of the metadata schema

23 Oct 2018

Dear all,
as we discussed in last week's meeting, we need to finalize the first
draft of the metadata schema before the end of this week. Thanks to
Louise, we have a draft document that already takes shape, see [1].
We also have some discussions, but I guess we still could use some
more feedback. So, please have a look and share your thoughts.
Best regards,
Rolf
[1]: https://docs.google.com/document/d/1wxSwzvpBB96a0k168NEqS3ZFuPBtSfEcXDwJ...
--
Rolf Krahl <***@***.***-berlin.de>
Helmholtz-Zentrum Berlin für Materialien und Energie (HZB)
Albert-Einstein-Str. 15, 12489 Berlin
Tel.: +49 30 8062 12122

  • Rolf Krahl's picture

    Author: Rolf Krahl

    Date: 25 Oct, 2018

    Dear all,
    I'm delighted by the many feedbacks and discussions on the draft
    metadata schema in the Google doc. Now P12 is approaching fast and I
    need to prepare the slides for the presentation of the first draft of
    the schema in our WG session. Unfortunatly, I need to make a cut now,
    not to stop the discussion, but in order to have some well defined
    state to base the presentation on.
    So I moved what I percieved as the current state of the discussion
    back to our GitHub repository, here:
    https://github.com/rdawg-pidinst/schema
    See the schema.rst in that repository. I took the opportunity to
    tighten and to rephrase some formulations and also to review the
    presentation, but I tried not to introduce any changes to the content.
    Due to lack of time, I didn't capture the explanatory text. This
    certainly still needs to be added (after P12).
    Of course, the draft schema is just that, a draft. It is still
    subject to discussions and modifications. In order to formalize the
    further procedure a little bit and make it easier to track changes, I
    suggest that new proposals for substantial modifications should be
    submitted as an issue to GitHub from now on. This would allow us to
    focus the discussion on this proposed change on this issue's page.
    Best regards,
    Rolf
    --
    Rolf Krahl <***@***.***-berlin.de>
    Helmholtz-Zentrum Berlin für Materialien und Energie (HZB)
    Albert-Einstein-Str. 15, 12489 Berlin
    Tel.: +49 30 8062 12122

  • Ulrich Schwardmann's picture

    Author: Ulrich Schwardmann

    Date: 21 Feb, 2019

    Dear Rolf, all
    yesterday one topic came up in my mind after Maggies question about the
    genesis of PIDs for instruments starting at a manufacturer and ending in
    a sequence of temporary owners/users of the instrument: what about
    instrument models or series?
    This would not solve the chaos induced by a sequence of temporary
    owners, but it would give the manufacturers at the very beginning a more
    feasible approach to deal with instrument identification from my point
    of view. This would help in the adoption process at the manufactures side.
    Also this way one would be able to identify classes of instruments with
    similar behaviour, but also to allow manufacturers to describe a model
    or series of instruments in a more precise way. This could be done
    possibly in a similar way as with the instruments itsself and should
    lead to identification of the model or series by an identifier. One
    could get access to documentation, firmware releases and default
    calibrations in a standardized way via the Identifier of the individual
    instrument, which would open a wide range of automated search and
    exploration possibilities with mayor impact.
    The current possible properties in the schema would be InstrumentType
    and RelatedIdentifier, but those are not specific enough for this goal.
    InstrumentType could also be 'microscope' or 'seismic station' and
    RelatedIdentifier could refer to a project or the manufacturer or
    whatever. One would not get the needed information by often misleading
    entries.
    My suggestion would be to introduce an additional property 'model' that
    describes the model or series of an instrument, in free text the model
    or series number given by the manufacturer and optional a PID referring
    to the model, also given by the manufacturer.
    This field, so at least the model or series number, should be mandatory,
    because the vast mayority of instruments are made by industrial serial
    processes and have such a series number. All others are probably build
    out of industrially produced sensors, where again the model or series
    plays a role.
    And also I would suggest to change a bit the descriptions of the
    properties in the given metadata schema to allow manufacturers to use
    the same schema for models or series.
    What do you think?

submit a comment