Attendees
---------
Brian Carrol -
Serena
Mary Ruddy -
SocialPhysics/Parity
Greg Byrd -
UNC/IBM
BrianC, MaryR,
Greg, PaulT,
TonyN,
Paul Trevithick -
SocialPhysics/Parity
Tony Nadlin - IBM
Mike McIntosh -
IBM
Valary Kokhan -
Parity Ukraine
Nataraj
Nagaratnam - IBM
David
Kuehr-Mclaren - IBM
Alex Amies - IBM
Jim Sermersheim -
Novell
Abhi Schelat - IBM
Agenda
======
1. Mary: IP
status update
2. Reminder about
the XRI IP
call
3. Next F2F in Austin
4. Status updates
by
component owners
a. HBX &
RPPS
(Abhi/Maxim)
b. i-card
manager (Paul
for SergeyY)
c. Token
Service (MikeM)
esp. builds, doc, etc.
d. IdAS
registry status
(Jim)
e. New Context
Providers:
file, mapping/policy 5. Service interfaces (esp IdAS) 6. OSIS-interop
scenarios
see [1] (esp. OpenID) 7. RP Enablement: Higgins project scope
discussion.
Relationship to
Pamela
project, Heraldry's dissolution.
[1] http://osis.netmesh.org/wiki/Interop_Use_Cases
Discussion
==========
1. IP status
update
-------------------
Mary: I reviewed
all open
items with Eclipse Legal yesterday. A total of 4 Higgins CQ’s are now
fully approved. Some items need the actual source, or we need to
revisit the
version number we want reviewed. Now that EclipseCON is over, Legal
plans a
review for the accelerated pre-approval for incubator projects.
Mary: Thanks
again to Higgins
presenters at EclipseCON. There was interest from other projects for
using
Higgins (including OHF and Corona)
and [of course] having Higgins provide the sign-in solution for RCP.
EclipseCon 2007
---------------
- Higgins/ALF
interoperation
demonstration by Brian and Mike
- Paul gave a
short talk
overview of Higgins
- Mary gave a
talk on open
source innovation
- There were some
good
discussion with MikeM
- Mary and Paul
talked to
the Eclipse OHF project
2. XRI IP call
today at 5pm
ET (details on the list)
----------------------------------------------------
- 2PM PT/ 4PM CT
/ 5PM ET on
Thursday
- Dial in:
1-641-297-5600,
passcode 762425
3. F2F will now
be the week
of 4/30 in Austin
---------------------------------------------
- Raj: We'll want
a session
on IdAS service interfaces
4a HBX & RPPS
Status
--------------------
Paul mentioned
that on
Monday Abhi's team was able to commit the "Zurich" stream of HBX and RPPS and
that
now Parity Ukraine would try to merge these into a single HBX.
Tony questioned
whether we
really can or should try to merge all features into a single HBX. Paul
mentioned that this certainly was his hope.
There was a
discussion that
we agreed we'd continue at the F2F about balancing complexity
management with
making this easy for the user (the paramount consideration), etc.
Paul: One more
item. We've
started a dialog with Kevin Miller the developer of the Firefox
Perpetual
Motion add-on. He's looking into EPL licensing and would like to
collaborate
with us.
4b I-Card Manager
Status
------------------------
Paul mentioned
that Sergey
Yakovlev (Parity Ukraine) is working on a new I-Card Manager UI and
that work
should begin to surface in a couple of weeks.
4c: Token Service
Status
------------------------
Paul mentioned
that he knew
that the Eclipse ALF project (Brian), and ooTao (Andy Dale) have been
able to
build and deploy the token service. Mike mentioned that Praveen
<unknown>
from AOL has too.
Paul: we are
working on
building automated build scripts
Mike: that would
be very
beneficial. We're getting more and more folks asking questions, etc.
and
servicing the interrupts is taking a lot of cycles
Mike: the simple
XML-based
context provider isn't writable right now.
Valery: you could
use the Jena
based provider. I
don't think it is too difficult to use.
Mike: we have a
de facto
profile for how we interact with the context provider by the STS. I
haven't
even tried to use the Jena
one. Valery, have you documented the steps to deploy the Context
Provider?
Valery: not yet.
Mike: I can try
it but my
guess is that it has to be customized in order to support the attribute
set
required by the TS. We need to figure out those details. I'm stuck with
using
the ones I know how to configure.
Valery: I will
try to document
how to work with the Jena-based provider. When you create a context it
requires
a schema.
Mike mentioned
that his goal
is to make his code as CP-independent as possible and that the
differences in
how they are configured makes this difficult.
4d: IdAS registry
status
------------------------
Paul asked about
the status
on implementation of the new registry design?
Jim said that
he'd been
working through a wave of refactoring based on the update changes (and
also
some java 1.4 changes) first and hadn't yet gotten to it and was
looking for
volunteers.
Paul mentioned
that Valery
had some code to contribute to this effort.
4e: New CPs
-----------
There was a
question about
the motivation behind the new "mapping" (aka policy) Context
Providers that have been mentioned on the list recentsly.
Jim: the current
JNDI
provider has a bunch of LDAP-isms in it. They are not Higgins URIs that
we
need. So the JNDI provider needed a way to turn "mail" into something
appropriate. We started to put the code in but then we realized that
other
Context Providers that are serving up different kinds of back ends too.
Eg. in
LDAP everything looks like a dn. We decided rather than have a lot of
different
providers with different configurations, etc. It would be nice if we
had one
general purpose CP that consumed a CP. It can be set up in front of any
CP and
the mappings put it. This could give a nice consistent way to do this
mapping.
The reason the CP name has the word policy in it is that every method
runs
through this thing. You could do much more than mapping (e.g. access
control,
etc.).
5. Service
interfaces
--------------------
Tony expressed
that we need
to examine the Higgins APIs from a service interfaces point of view and
begin
considering the service descriptions as primary and the language
bindings as
derivative instead of the other way around. For example, we seem to be
putting
details into the APIs that probably can't be put into the service
interfaces
(e.g. for IdAS). He mentioned that David Kueher-McLaren and Alex Amies
have
been looking into the gap between the Java APIs we have and what might
work
with web services interfaces. David said he'd been looking at IBMs use
cases
and working with Alex on this.
6. OSIS Interop
---------------
Paul mentioned
about the
work at OSIS underway planning an industry wide interop event at
Catalyst in
June. See [1] above.
He pointed folks
to this
link and mentioned that the three separate Higgins/OpenID interop
scenarios are
buried in this page.
7. Higgins Scope
WRT relying
party code
---------------------------------------
There was
agreement that it
*is* within Higgins scope to develop "Relying Party Code" and that
this code should be multi-protocol (e.g. not limited to just CardSpace
support). We will work on interoperability with the Pamela project. We
will be
creating our own code. We'll reach out to Pam Dingle and try to
collaborate as
much as possible with her.
Misc
----
Brian: We're
consuming STS,
IdAS, and the JNDI provider. To date we've been building from the HEAD,
but it
might be time to tag stable builds.
Mike: nightly
builds are a
pre-requisite.
Paul: Parity Ukraine
is
working on those. Hopefully done in a couple of weeks from now.
Abhi:
Announcement: the IP
issues related to the idemix code should resolved by tomorrow.
Action items
============
* Work on agenda
for F2F
* Logistics for
new date for
F2F
* Valery to
document Jena CP
deployment/configuration