Child pages
  • ULX Teleconference 2010-09-27
Skip to end of metadata
Go to start of metadata


  • Time: 08:00 PDT | 11:00 EDT | 15:00 UTC/GMT | 17:00 CEST (Time Chart)
  • Skype: +9900827042954214
  • US Dial-In: +1-201-793-9022
  • Room Code: 295-4214


1) Roll Call
  1. Paul Trevithick *
  2. Axel Nennker *
  3. Scott Cantor
  4. Keith Uber *
  5. Philippe Clement
  6. Gael Gourmelen
  7. Bob Morgan
  8. Benoit Bailleux


  1. John Bradley

*Voting members

This meeting was quorate (3 of 4).

2) Minutes

We approved the following minutes:

3) Nominations for co-chairs

Scott: Made the motion that Bob Morgan and Paul Trevithick continue as co-chairs
Gael: Seconded
...the vote passed unanimously thus Bob, Philippe and Paul are now the co-chairs

4) OASIS UI Profile for SAML Metadata

Paul: What does "containing role" mean?
Scott: SAML metadata is organized into entities and roles. And entity is a network service. Different roles are defined for different kinds of roles. IdP and SP are the principle roles. Extensions are expressed at the entity or a the role level. E.g. an entity that was acting as an IdP and an SP role.
Paul: Why are attributes not mentioned?
Scott: Because they are mentioned elsewhere ?(i.e. in other SAML specs). SAML's history is one of more pre-configured relationships vs. more dynamism as with Infocard.
Scott: We did reject having any alt tag for the icons. Also updated the spec that display names should be amenable to accessibility.
Keith: Is there support for localizing the graphics for each language?
Scott: Yes, that is supported
Bob: A distinction is being make between elements in here that are signable vs. things that are okay to be self-asserted. Looking at the URL statements (e.g. privacy statements), their content is variable.
Scott: Some parts of this spec were inputs from existing federation deployments. Some is overkill from my point of view.
Bob: We might want a schema that is independent of the kind of document. If you want to stick the element in metadata that's okay, if you want it in an XRD that's okay.
Scott: This is problematic because XRD has evolved to simple property-based vs. XML extensions. We might want an abstract model above the level of these two expressions.

5) Paris ULX meeting

Proposed agenda topics:

  1. Next documents to publish from ULX WG
  2. Extension of the "driven-by-RP" UX scenario to an "ISA based" scenario: Is there any additional work to do
  3. Claims & Metadata: What are they, and how to circulate them among the actors (RP, IdP, ISA)
  4. Other suggestions?

Scott: the problem with the ISA model is that the ISA knows a lot less about the RP/SP. We've looked at it this way: if you want to get the ISA involved then we assume it has access to metadata, and if it wanted to get some more RP/SP-specific informaiton (e.g. a filtered list of IdPs) it could fetch the XRD from the RP/SP.
Paul: It appears that coming up with the RP/SP XRD is the next logical piece of work. Perhaps we could work on this in Paris.

Next Teleconference
  • Date: Monday, October 4, 2010
  • Time: 08:00 PDT | 11:00 EDT | 15:00 UTC/GMT | 17:00 CEST
  • Dial-In: Skype: +9900827044630914, US Dial-In: +1-201-793-9022
  • Room Code: 295-4214
  • No labels