Attendees:


Regrets:

Pre-Reading

Agenda

  1. Open Actions
  2. Administrivia
    1. ldapPublicKey
    2. Shifting from the Work Item Portfolio to GitHub issues
      1. how will that work for general issues such as requests to take in new (old) work?
  3. Experimental attributes
  4. MACE SAML Attribute Profile (is this the correct version?)
  5. Subcommittee status
    1. voPerson
  6. Status of splitting the eduPerson spec into core and supplemental
    1. Splitting Protocols From Schema Documents


Notes

  1. Open Actions
  2. Administrivia
    1. ldapPublicKey - Still in progress; Heather to follow up with Nicole
    2. Shifting from the Work Item Portfolio to GitHub issues
      1. how will that work for general issues such as requests to take in new (old) work?
      2. We could consider using GitHub Projects to sort and manage issues (including requests for new attributes, bugs, etc). The AAF uses GitHub for this, creating a "Stockpile" repository that they use to collect everything, tag the issues for organization, etc.
      3. SEB will shift to using GitHub instead of the wiki to track areas of work
  3. Experimental attributes
    1. On the one hand, this feels like requiring multiple implementations of the same thing as the information must ultimately change when the item comes out of experimental. Alternatively, this could allow for us to support prototyping and encourage proof of usefulness. Managed schema shouldn't be a dumping ground for anything and everything.

    2. As per today's discussion, here's a proposal instead of creating a new Experimental status. All new attribute proposals for SEB-managed schemas would go through the following process:

      (1) Proposers create draft (possibly in the form of a pull request).

      (2) SEB performs quick sanity check for relevancy/interest.

      (3) OID assigned by SEB to each new attribute.

      (4) Draft reviewed by community (open comment period).

      (5) Proposers present a relevant Proof of Concept using the new attribute(s), and incorporating any community feedback resulting in changes to proposal.

      (6) SEB performs final review/approval.

      (7) Draft merged to next version of the appropriate schema.


  4. MACE SAML Attribute Profile (is this the correct version?)
    1. This document defined, pre SAML2, how to represent eduPerson attributes with a friendly name. The hope is to publish this in a controlled space. Heather will handle it.
  5. Subcommittee status
    1. voPerson - no change
      1. Note that the schema is heavily used in the research community, but the mechanical bits are what's taking time. Suggest move resolving the open issues (which are largely formatting ) to 2.1, and go ahead and migrate voPerson to 2.0 under SEB management
      2. Heather will send out an e-vote to the SEB asking them to look at the open issues, decide if there are any showstoppers towards going ahead and bringing this in as a 2.0
  6. Status of splitting the eduPerson spec into core and supplemental
    1. Splitting Protocols From Schema Documents
    2. Schema part has been split (see https://github.internet2.edu/api-schema/attribute-dictionary/blob/draft/eduperson-entries.adoc). Still need to go back to the eduPerson spec itself and figure out which parts still need to be in there that haven't been covered.
    3. Request that the SEB look at a new scheme - something between a natural language attribute dictionary and a data model, called JSON abstract data notation. In theory, this would allow for automatic processing into any other desired representational form.
      1. is this a necessary and/or useful layer of abstraction? It seems like a lot of work for a schema that is not that complicated.