Re: Tentative OPES Agenda for IETF 53


There are some very good protocol issues in the personalization drafts.

I'm not certain that they change the architecture, though.  For example,
all the content path protocols already have a notion of user session,
and an intermediary can understand this and apply personalization
rules without any architecture extensions.  More complicated notions
of session, based on exotic cookies, for example, might have to be
programmed into the intermediary, but the identification of the session
can be communicated to the callout server through context variables
(most likely header extension fields).

For better or worse, the original architecture relegated all issues
regarding things like user profiles to secure configuration to be
protected by AAA stuff.

A question for discussion is whether or not P3P is an appropriate
model for the chained trust relationships that intermediaries need.
On the other hand, I've borrowed heavily from the W3C work on signed
content in developing the data integrity mechanisms that OPES could
use.  That seems like a good model for developing the user profile
syntax and protections mechanisms.

We'll need a good, long-term liason if we are to align with W3C.  I'd
not have guessed that "multi-modal" would have anything to do with
OPES, for example.


Abbie Barbir wrote:

Hi all,

In my opinion, the work on personalization directly line up with atleast the original OPES work.

I also strongly belive that the OPES senarios should also align with thework that is currently done in W3C in multi-modal working group.

There is also, correlation with the W3C webservices initiative, at leastfrom the security prospectives.

The above should be taken into consideration when we look at themodel/secnarios document.

cheers

abbie

    -----Original Message-----
    From: John G. Waclawsky [mailto:jgw(_at_)cisco(_dot_)com]
    Sent: Tuesday, March 12, 2002 12:46 PM
    To: Condry, Michael W.
    Cc: Knight, Paul [BL60:1A00:EXCH]; OPES Group
    Subject: Re: Tentative OPES Agenda for IETF 53

    Could you expand a little on why you do not consider personalization
    an OPES item? At a minimum I believe it is strongly related because
    opes could provide new ways of accomplishing it.    Regards  John

    Condry, Michael W. wrote:

    IMHO Personalization is an interesting topic area but not an OPES
    item in my opinion. Clearly I was interested in this area!

    It would be useful to extract the key observations made and
    discuss them on the mailing list/working group.

    At 06:34 AM 3/12/2002, Paul Knight wrote:

    Hi,

    Although I agree the workgroup charter and IAB considerations are
    obviously the primary focus of our work, I did want to point out
    another two (non-WG) IDs which contain valuable input to the
    discussions:

    - draft-barbir-opes-spcs-01.txt, "Requirements for an OPES
    Service Personalization Callout Server", which contains some good
    discussion of callout server issues

    - draft-barbir-opes-fsp-01.txt, "A Framework for Service
    Personalization". In addition to personalization elements, this
    also contains a useful section on threat analysis and security
    mechanisms, which may be a good starting point for these issues
    across OPES.

    I hope these will be useful as input to the WG discussions.

    Regards,
    Paul

    -----Original Message-----
    From: Markus Hofmann [ mailto:markus(_at_)mhof(_dot_)com ]
    Sent: Tuesday, March 12, 2002 1:54 AM
    To: OPES Group
    Cc: Hilarie Orman; Allison Mankin; Ned Freed; Patrik Fältström
    Subject: Tentative OPES Agenda for IETF 53


    Hi,

    below the tentative agenda for the OPES meeting at IETF 53. Focus
    will
    be on the next immediate document deliverables and on how we want to
    get there - and less on existing Internet Drafts that have
    previously
    been submitted with reference to OPES.

    This also implies that we do not necessarily take existing IDs as
    starting points for the WG docments, but that we rather consider
    them
    valuable input, taking into account the final charter and the IAB
    considerations for OPES.

    It is important that we get more discussions going on the list prior
    to the meeting. Thanks to Hilarie for initiating a threat on content
    path security/encryption. Please read Hilarie's posting and
    comment on
    it. Also feel free to comment on relevant issues listed below.
    Without
    discussion, no progress!

    Cheers,
       Markus

    =============================================================

    Agenda for OPES Meeting at IETF53
    =================================

    - Introduction, minutes taker, blue sheets
    - Agenda bashing
    - Charter walk-hhrough
    - Discussion of workplan, milestones, and how we want to get there.
    - Disussion on the end-to-end integrity and encryption compatibility
       issue for proposal to the ADs.
    - Discussion of next WG documents
         - Scenario document
         - Architecture document
         - Callout protocol/tracing requirements
         - Endpoint authorization and enforcement document
         - Threat/risk model document

    Related Internet Drafts (non-WG drafts)
    - draft-barbir-opes-vpcn-00.txt
    - draft-dracinschi-opes-callout-requirements-00.txt
    - draft-elson-icap-00.txt (expired)
    - draft-mchenry-opes-deployment-scenarios-00.txt (expired)
    - draft-rafalow-opes-policy-requirements-00.txt (expired)
    - draft-tomlinson-opes-model-01.txt


    Michael W. Condry
    Director,  Network Edge Technology