Present: Alex Stuart and Albert Wu 

A small WG this week so we had a relatively wide-ranging conversation:

  • Main work over the next few weeks is to work out what we are going to say at TNC. This WG has a good shared understanding of the problem space, the history of how we got here, and the 2 proposals (select a profile defined elsewhere; put suitably-encoded trustinfo metadata into an entity attribute). We need to condense that into a short presentation at REFEDS. And we need to explain that we plan to have two parts to this work
    • Part 1: define the entity attribute for profile
    • Part 2: more general case, a container for full trustinfo metadata, perhaps for IdPs as well as SPs. We need some concrete use cases to determine whether it's worthwhile to develop the spec.
  • Albert won't be at the meetings on 16 an 23 May (although will be working); Alex won't be available for the 23 May)
  • Alex needs to write a REFEDS blog post about WG progress. We must ask people "bring your use cases"
  • JSON spec seems less developed than XML (for example whether one has arbitrary set operations, or combination of include/exclude statements)
  • XML spec has some ambiguities: most of the queries assume some data structure (e.g. registrationAuthority has counterpart in SAML). However md_source appears to be a SeamlessAccess implementation detail
  • NB: 2 parts to trustinfo metadata
    • filtering
    • adding entities
  • be careful - we don't want to define a full query language
  • Albert educated Alex about trustmarks in OID federation. There's a distinction between the trustmark issuer and the mark itself. Which we noted was good because REFEDS has experience of defining standards which do not have unambiuous interpretation across federations (looking at R&S in particular)
  • No labels