[DG-AM] Notes from breakout session regarding attributes from the Mapping the Identity Ecosystem workshop
hlflanagan at internet2.edu
Tue Dec 20 14:14:49 EST 2011
Who is interested in attributes? - Notes from the ISOC Identity Ecosystem Working Group Open Groups defining standards / schemas / frameworks
OIX Attribute Exchange: only having its first meeting this week .
MACE-DIR / eduperson / SHAC.
More closed defining standards / schemas / frameworks
ITU SG 17 U10: open to liaison with experts on board and open to members.
These are possibly at a higher level than attribute schema.
Discussion / implementation groups:
Kantara Attribute Discussion Group.
REFEDS Attribute Release.
Sharing across the groups
it is difficult to share across where information is closed off by non-release before final / copyright / behind closed doors. The earlier in the process you share work, the easier interoperability is later. Where is convergence?
How are we ensuring that the correct feedback is being created when outputs / schemas / proposals are coming out of these groups? Is Kantara the feature for this?
making John Bradley be the lynch-pin.
use existing discussion groups – Kantara is the obvious place.
Go out to all the places having discussions – unscaleable.
What is the “internet of things” attribute schema?
Attributes of things as well as attributes of people.
Devices – as identity consumers as well as having identities themselves. Is device identity more a projection of personal identity?
Too much, too many, too patchy.
Collision-resistant namespaces - the OpenID Connect approach.
A simple way to design attribute schema.
Things that don’t assume that there will be one of every type.
Agreement around extensibility.
Context provider – in what context can I use this context?
Where does the conversation about this happen? Who operates an attribute registry? We need real participants who will benefit from this.
Developing the business case, putting time and effort in to adoption of what we have now.
Discussions to be had around what is the right scope to put around the structure of attributes – filtering, flexibility, assertion.
Language code – still not a coherent way to express preferred language.
write up an RFC on problematic points?
OVERALL ACTION: Give these notes to OIX and Kantara
More information about the DG-AM