This is a very important announcement:

I will do all the TMG CONTENT Stuff in future.

http://conferencing.sap.com//global/SyncEvents/EnterEvent.asp?campus=global&campusURL=conferencing.sap.com&Lang=EN&external=627008

Agenda

1. - Welcome
2. - Agree on agenda
3. - Select note taker
4. - Review Open Action Items
5. - Discuss possible comments to Tonys outline (see document in UCM ODP 3)
6. \ - Discuss possible comments to drafted definitions (see document in UCM ODP 3)
7. \ - Discuss further work/ agenda for next meetings
8. - Other business
9. - Next meeting

Call details

EVENT ID: 627008

FIRST TIME USER: Please login 5 minutes before the meeting starts.
Online meeting > connect with PC first then with audio device!
This invitation is only for you, please do not forward to others.

---------- NOTICE TO ALL PARTICIPANTS --------------------------
Please note that any recording of the meeting is illegal unless all participants have given their prior consent. Accordingly, the recording function has been disabled by default. It can only be reactivated by the organizer of the meeting, who must then ensure that all participants will be notified thereof prior to the meeting, or at its beginning, at the latest. Any objection to such recording by a participant must be duly followed.  

---------- SAP Participants ------------------------------------
JOIN EVENT:
Click JOIN Button in Outlook Appointment if available or click:
http://conferencing.sap.com//global/SyncEvents/EnterEvent.asp?campus=global&campusURL=conferencing.sap.com&Lang=EN&external=627008

AUDIO:
The use of a PC headset instead of telephone dial-in is recommended!

TELEPHONE DIAL-IN NUMBERS:

• SAP Germany (internal):  49-6227-7-55555
• SAP US (internal):       1-610-661-4000
• Europe (most countries): 00-800-4545-7878
• US:                      1-866-893-3308
  Full list of dial-in numbers: http://conferencing.sap.com/global/default.asp?target=dialInInformation

TELEPHONE USER-ID / PASSWORD:

• Telephone User-ID : SAP User-ID  > C=2, D=3, I=4 (e.g. D012345 = 3012345)
• Password: Set 4-digit password once prior to your first event
    > https://conferencing.sap.com/global/ChangeTap
    > Note: Enter SAP_ALL User ID and SAP_ALL Password

---------- Non SAP Participants --------------------------------
JOIN EVENT:
http://conferencing.sap.com//global/SyncEvents/EnterEvent.asp?campus=global&campusURL=conferencing.sap.com&Lang=EN&external=627008

TELEPHONE DIAL-IN NUMBERS:

• Europe (most countries):  00-800-4545-7878
• US:                       1-866-893-3308
  Full list of dial-in numbers: http://conferencing.sap.com/global/default.asp?target=dialInInformation

TELEPHONE USER-ID / PASSWORD:

Note Taker	Øyvind Aassve
Participants	Anthony B. Coates, Gunther Stuhec, Alexander Schmidt, Rob Chumbley, Andrew Hearn, Nada Reinprecht, Anders W. Tell

Agenda

1. - Welcome
2. - Introduction to the work of the group w/Tony
3. - Propose outline for the draft document. Structure, fit in requirements.
4. - Definitions of context and context model
5. - Relationship to Metapatterns
6. - Discussion on contributions and central concepts of the UCM-specificaton (SAP and MileyWatts)
7. - Brief meeting with TBG 14 regarding context categories.
8. - Next meeting

Meeting Results

Introduction w/ Tony:

See first 5 pages of Tonys Contribution:
http://www.unstandards.org:8080/download/attachments/3801826/UCM+Context+Model+and+Methodology+Proposal+-+Coates.pdf?version=3

Goals - automatic processing and search (what components are existing). Some associations might also just be for making it easier to for people to understand - not only for the machines.
Context is in the eye of the beholder. There might be different opionons given context of something. There is an idea of incompleteness. Nothing you will do will ever be complete. Build for an open world.
Time Constraints: Recommended article of Martin Fowler - "Things that change in time".
Comes down to putting date (start - end (or open)) on things.
Validity - recognize that there can be variable qualities of descritions. Add things as wrong without taking out straight away.
Skepticism of the correctness of everything. Healthy.

Outline of specification:

Tony has prepared. No immediate questions to Tonys outline and requirements classification. Will be taken up on next meeting after people have had time to look through them.

Definitons:

Context Defintions (see proposal in Tonys contribution):

Activity - overloaded word. Proposed alternatives - environment, things.
Take out second part of first line?
Categorization - should be put in - important use of it. But again only one of the possible uses.

Context Metamodel:

Second line - Anders - proposal to include vocabulary. Glossary - what does the different words mean.

Relationship Metapatterns

Peter Wisse's papers have some very interesting concepts and he has been working with context over a long period of time. These includes open world assumptions, contexts dependence on view, awareness of variations in quality of context descriptions, and the time dimensions influence on context. Some of these concepts will be useful for the UCM methodology. However, our work is limited to UCM and it is not possible for us to discuss changes to the concepts of CCTS as proposed in some of his feedback.

Peter Wisse and Jan van Til are added to the yahoogroup and invited t0 future meetings.

Contributions:

Tonys metamodel:

Comments:
There should be a statement in the beginning noting the importance of that this is one of the most important specifications for B2B collaborations.

Gunther - is intersection/ exclusion is the best solution. Will use
Boolean expressions which can lead to very many and/ors.
Context association- connection between two contexts. Important to know whether Association Type is transitive or not with regards to automatic processing.
Mathematic logic - collections and description logic is important.
Narrow/ broader makes the math more difficult. But we cannot escape this, necessary to take into account the necessary math for sufficient logical expressions.
Primary categories: Tony's suggestions - to define the minimum amount we can get away with.

Anders - types of associations? - - main set of context associations is an important part of the specifications (part-of is identical to broader-narrower), same as, disjoint, inverse, broader. Add associations that can be useful - not because we can think of it.

Time is not category - only for lifecycle of other categories. Valid to and from.
Model live over time, certain things will come and go. Alternative to versioning. Can issue versions every time something changes. With dates you can say at any date what is the model, so that you know from this date you should not use this part. Application can read this - drop pre-from and past-to.
But if you for every value you should define date to and from it can be very complex in maintenance. The only thing that is required is value from date.

Anders math - important with set of examples. Also for validating what we are doing.

Gunther: Challenge from CCTS - how do we do qualifier. Qualifiers are names. Semantics of information vs. Semantics of context. How do we define unions that creates new categories.
Narrower/ broader - replaced part-of - assumed to be more generic.

SAPs metamodel:

As SAPs work had been released just ahead of this meeting. Gunther asked if someone had an issue with that with regards to the IPR policy because material was still labelled with SAP logos. Anders Tell had, and therefore it was decided not to go forward with a discussion on SAPs contribution.

Furhter work:

How to deal with several submissions. Suggestion that we work with a mixed model where we accept wholistic contributions and also work out position papers that addresses possible gaps, missing parts in these contributions. Anders informed that OMG only takes full contributions and continue working on several paralell specifications. Allows you to follow different trains of thoughts.

Meeting with TBG 14

We had a brief discussion with a few central members from TBG 15 (Nikolaus Sahling, Ed Buchinski, Collin Parker) with regards to a previous action point from. The purpose from UCMs point of view as to get a confirmation that TBG 14 that they can provide a set of context categories with corresponding context values.

TBG 14 presented their under development framework for categorizing business process and lifecycle-processes. These are clear potential candidates for category classifications. However, the context methodology will need several more context categories in order to provide a fullfledged context description.

The conclusion was that UCM must establish contact on a higher level with TBG because we need a broader level of context categories. Also, TBG 14 recommended that UCM work out a template for what kind of format and information we want the TBGs to provide.

Action Items

Gunther - cleans contribution from SAP-logos.
Oyvind adds Peter, Martin and Jan to Yahoogroup.

Next meeting