[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
[higgins-dev] Notes from Higgins-dev call 3/1/2007
|
Attendees
---------
Mary, Mike, Paula, Andy, Tom, Duane, Dale, Daniel, Jim, Abhi, George, Brian,
Valery, Paul, Tony
Agenda
------
1. Mary: IP Update [1]
2. Paul: Milestone 0.8 Review [2]
3. Paul: OpenID interop approaches
4. Paul: IdAS Registry [4] <--never got to it
[1] http://wiki.eclipse.org/index.php/Higgins_Third_Party_Dependencies
[2] http://wiki.eclipse.org/index.php/Milestone_0.8 (updated per meeting)
[4] http://wiki.eclipse.org/index.php/IdAS_Registry_Proposal_2
Abhi: Will post a 12mb movie showing interface to YouTube (see action items)
IP Update
---------
Mary presented and we discussed the dependencies "status" page [1] above.
There were questions about the semantics of the columns. The explanation has
been added to the wiki page.
Abhi: Eclipse has approved IDEMIX but lately there are some new internal IBM
issues.
Mary and Abhi to follow-up on status.
Milestone 0.8 Plan review
-------------------------
We walked thru [2] above and made some changes and additions as we went
along.
- Novell has started work on an ISS Client UI in C#
- We discussed the pressing need for an easy-to-configure CP (preferably
file-based) that also supports update operations (not just read-only)
- This is what's holding up documenting the TS build/configure
- The Jena cp would be easy but it can't be used due to IdASRegistry
integration issues
Paul: WRT I-Card manager we present to the user notions of "groups" and
"i-cards". Now every i-card represents exactly one Digital Subject. A group
is a set of i-cards (e.g. a directory, social network, buddy list, etc.). We
wrote http://en.wikipedia.org/wiki/I-card.
There was a long discussion mostly with Mike, Jim, and Dale and others about
the fact that in IdAS you authenticate to an entire IContext when you open
it, there is no finer-grained authentication (e.g. to a specific DS). With
this approach the threat model is highly dependent on (i) Context Provider
implementations (ii) deployment/configuration issues especially related to
"system" vs. "user" authentications (iii) deployment architecture: can an
attacker go behind a Context Providers, back, etc. We didn't come to any
hard and fast conclusions. Just ventilated the issues.
We never got to agenda items #3 or #4 above.
Action items
============
- Abhi will post UI video to YouTube
http://www.youtube.com/watch?v=Nuv7T3E306c (still being processed.)
- Mary and Abhi to follow-up on status of Idemix IPzilla item
- Valery to check with Maxim on combining HBX functions in one HBX