Meeting Notes:
POSC Member General Meeting
June 1998

Note: These notes are subject to review. Members are welcome to send comments to Alan Doniger at Doniger@POSC.org. Some presentation slide files are linked into the notes. Additional links to presentations will be added as they are supplied. Notable changes to the meeting notes will be summarized here.

These meeting notes supplement the meeting materials distributed to attendees. Issues and action items are highlighted.

Questions about POSC in general or about these meeting notes can be directed to Alan Doniger (Doniger@POSC.org) or Paul Maton (Maton@POSC.org).

Strategy Sessions

• C. POSC Board Report; Bob Pindell (POSC Board), David Archer (POSC)
• D. POSC Maintenance Work Category Plans ; Jim Brannigan (POSC)

Project Report Sessions

• E. FieldBank Project: Gravity and Magnetic Extensions to Epicentre; Kitty Hall (Ark Geophysics)
• F. WIME Update; David Roper (Sysdrill)
• G. OpenSpirit Alliance Update; Nico de Rooij (Shell)

• H. RESCUE Alliance Update; Knut Kirkemo (Statoil)
• I. URGENT Update; Sheena Henderson (PDS)
• J. Shell/Oman Project Results; Gina Severeijns (Shell/PDO) and  Floris Versteeg (Oracle)

Project Proposal Sessions

• K. Interoperability and Business Objects Request for Technology (RFT) Process Update; Alan Doniger (POSC)
• L. POSC Shared Earth Model Project; Dan Schenck (POSC)
• M. Coring and Core Analysis Project Proposal; Bob Skopek (Petrophysical Applications International)
• N. Vendor Data Store Mappings Project Proposal; Steve Trythall (PrismTech)

Closing Sessions

• W. Facilitator's Discretion; Jenny Meader (POSC Consulting Ltd.)
• X. Topics for Future Meetings
• Y. Action Items

Strategy Sessions

C. POSC Board Report; Bob Pindell (POSC Board), David Archer (POSC)

(PowerPoint Presentation )

Bob reported on the results of the June 15 POSC Board meeting. At that meeting, the Board approved the revised POSC by-laws, determined how the Board will transition to the new member-elected form, discussed pending election for four Board seats, and reviewed the status of the business plan. Bob reiterated the main driver for the recent changes in POSC's governance and funding models, i.e. to evolve according to the original vision of POSC as a perpetual, sustainable, membership organization. This year, three sponsor directors (BP, Chevron, and Mobil) and one at-large director (Jean-Paul Marbeau) will transition off the board to be replaced by four elected Board directors: two representing oil companies, one representing non-oil companies, and one representing not-for-profit organizations. In '99, one sponsor director (Texaco) and the remaining two at-large directors (Bertrand du Castel, Keith Steele) will transition off the Board. In '00, the remaining three sponsor directors (Elf, Saudi Aramco, Statoil) will transition off the Board.

Questions and Answers:

• Jenny Meader: What is the term of office for elected Board directors? Answer: Three years.
• Steve Daum: Is the voting uniform for all Board directors? Answer: Yes.
• Stewart Robinson: How is an oil company defined in the medium term? Answer: It is easy to define what an oil company is now, i.e. an organization that primarily does business by taking equity interest in oil and gas properties. If suppliers begin to take equity interests in oil and gas properties, then we will have to look carefully at revising this definition. I believe that the definition will be based on the predominant revenue stream of the organization.
Next, Bob described the nomination process for the upcoming election. A Call for Nominations by organizational category will be issued in early July. Nominees will be validated for willingness and ability to serve. If more than one isnominated from the same company, POSC will consult to determine the one nominee that will stand for election from that organization. Then, ballots will be distributed to the entire POSC membership. Member organizations will have one vote per member organization for each open seat. The vote will be by written proxy or by paper ballot handed in at Annual meeting on October 20. The election results will be announced at Annual Meeting.

Questions and Answers:

• Steve Daum: Was consideration given to weighting the vote so no-oil companies do not dominate the voting for the oil company representatives? Answer: The effect of the voting rules was considered. The principal that led to universal voting was that all of the directors represent the entire industry rather than one organization or even one industry sector.
ACTION ITEM C.1: POSC will publish the list of member organization by election category as part of the call for nominations issued in July.

Bob reported that the POSC business plan for this year has been approved by the Board. A recent draft of the plan is posted on the POSC Web Site. The final version will be posted there soon. Bob summarized the actions to be taken by the Board in the near term: (a) to share the status of the Board's work, as is being done today, (b)  to develop nominees for the Board seats that are up for election, and (c) to listen to member feedback.

(David's Presentation; requested)

David began by expressing appreciation to Alain Robert for his service to the POSC Board since the inception of POSC. Alain left the Board as Elf's director recently to be replaced by Bernard Potier.

David explained the addition of a new lowest level in the member fee structure, i.e. $2000 per year for organizations up to 1/2 million dollars. This was done in response to member feedback.

David briefly reviewed the three POSC Work Categories and identified the primary individuals responsible for each: Cary Purdy for Work Category I, Alan Doniger for Work Category II, and Dan Schenck for Work Category III. David explained that there had been concern expressed by the POSC lawyers about the process for expanding the baseline of specifications and related material that is subject to maintenance support in Work Category I. He explained that there will be a process of 'voting' by the membership to approve such expansions of scope. He explained that POSC's fee-for-services business, i.e. Work Category III, will be pursued in cooperation with other service companies as opposed to in competition with them.

David explained that there is now a marketing plan in the business plan. POSC is recruiting a Business Development and Marketing Manager now. The key aspect of the marketing plan is face-to-face contact (as opposed to mass media campaigns). David referred to the brochures describing several ongoing and proposed POSC projects and describing the new Knowledge Management joint venture. He invited everyone to take and read a copy of the new InfoPOSC magazine.

Questions and Answers:
 

• Hamish Wilson: Product definition is key to a marketing plan. What is POSC selling? Answer: In short, POSC sells collaborative work that supports the POSC mission.
• Eric Hatleberg: The POSC Membership can generate ideas for things POSC should sell.
• David Archer: We will instigate work ideas and we expect the industry to come up with ideas, too. We sell fee for service / licensing work. This includes consulting, training courses (delivered by ourselves and third parties), project management, direct technical work, certifying providers (training, consulting), business object registration, etc. And, of course, we sell POSC memberships.
• Hamish Wilson: How much of the consulting services will compete with us (Paras)?
• David Archer: We will have to work carefully with you to differentiate the work we do from the work you do. We believe that there is an overabundance of demand and limited supply of qualified consultants.
• Keith Steele: There is a shortage of qualified POSC consultants.
• David Archer: We want POSC to be known as the place to come to find the right source to address a need and to broker the right partners and resources.
• Tim Haynes: Is the Knowledge Management Joint Project in Work Category 3? Answer: Yes.
• Tim Haynes: What is the status of the POSC (Europe) office in London? Answer by Michael Ring: We are negotiating for space in the Richmond area which will reduce our costs. The office will be on George Street above the post office. We will be able to host training classes and meetings on a limited basis. We expect to move in by the end of July.
• David Roper: Where is Richmond? Answer: At the end of one branch of the District Line of the London Underground. It is a seven-minute walk from the train and Underground stations.
• Tim Haynes: What are the plans for the long term? Answer: We are not sure of the right size for the London staff. It will depend on the quantity of business generated.
• Yvon Kroese: Does POSC have a recruitment plan? Answer: We are now looking for about six people, e.g. SEM, interoperability, reservoir engineer, marketing, etc. We will not hire people and then look for business for them to do.
• Steve Daum: We have been used to two-year staggered system for the at-large board directors. In the new plan, it appears as though the terms of the directors will be less revolving. Would you consider annual at-large director elections? Answer: There is a system of staggered rotation in the plan.
• Bob Pindell: All board members are at-large, representing the industry as a whole.
• Bertrand du Castel: The transition will maintain experienced board members while new members join the board.
• Bob Pindell: We are considering how to broaden at-large-like representation during the transition. Bertrand du Castel handles member communication on Board issues. We want to broaden communication on a regular basis.
• Valerie Calderbank: Does the POSC Certified Consultant idea remain vague. Answer: It is less vague than before. We are thinking about how to put it in place.
• Nico de Rooij: It is important for (potential certified) consultants to write down what they think it is? Answer: Good idea.
• David Roper: Will consultant certification be on a personal or organizational basis. Answer by Dan Schenck: We have talked about both.
• Bertrand du Castel: The previous strategic alliance proposal failed because POSC decided to certify non-members. This mistake should not be repeated.
• David Archer: Yes, POSC membership will be prerequisite to certification.
ACTION ITEM C.2:. Member organizations with an interest in becoming POSC Certified Consultants will send to POSC a description of how the certification process should be defined. POSC will request this input from member organizations.

D. POSC Maintenance Work Category Plans ; Jim Brannigan (POSC)

(PowerPoint Presentation)
(Handout Document; HTML)

Jim described the scope and processes associated with specification refinement, also known as maintenance or POSC's Work Category I. He explained the current Web document that describes a number of Member Choice maintenance work items for which member feedback is being requested. Based on the feedback from the membership, POSC will choose the items to work on and will organize them into plans for future specification releases. Jim walked through the eight proposed Member Choice work items.

Questions and Answers:

• Stewart Robinson: POSC is an organization that does not deliver commercial software products. I am dubious about the need to continuously change the specifications. I don't want a regular new version. It will cause me grief. If there is something that stops people from using Epicentre, it should be fixed. Answer: I agree. That is the point of the exercise. We want to limit the scope of change to the essential work items based on your input.
• Eric Hatleberg: If you look at the Member Choice work item list, the main areas deal with enhancing the Reference Data.
• Stewart Robinson: That's OK.
• Michael Ring: There is a fundamental message from software suppliers. They say, "If you don't expand the scope of  the data model, it will cease to exist."
• David Roper: Until the time that Epicentre fulfills its intended scope, it must expand.
• Stewart Robinson: I am not saying to freeze it. Does POSC have processes that will provide support more than one version of the specifications? Answer: We are working on processes to make small releases (i.e. patches) directed to a narrow area of need.
• Rick Morneau: Did you analyze what is keeping oil companies back from using POSC specifications or what is necessary as a driver to make companies want to implement Epicentre? How did you come up with the Member Choice list? What are the time and effort estimates on each work item? Answer: The Member Choice work item set was chosen so that in total they can be done within the budget for Work Category I Specifications Evolution. There are estimates for each work item.
• Eric Hatleberg: We looked at value of all candidate work items based on previously received membership input. We chose the items that will add value for both current and new users of the specifications.
• Dan Schenck: The Member Choice work items are based on member feedback about areas that need more work.
• <Conoco attendee>: Is there more detail on the Web site? Answer: Not yet. We will add more detail.
• Keith Steele: Concerning the work item to define a Java binding for the DAE specification. In a matter of days, a commercial DAEF product will be available. POSC should contact PrismTech in relation to this work. Answer by Dan Schenck: This is first exposure of this proposal to the membership. No work has been started on this work item.
• Keith Steele: We have ten people working in the area of the DAE specifications. Why would POSC want to divert resources to a DAE-related work item like this?
• John Gillespie: There are a limited number of POSC person-weeks. Knowing that number would allow the members to give better feedback. Answer: The number of POSC person-weeks available is not a limiting factor. Not all of the work on the proposed Member Choice work items will be done by the POSC staff.
• Jean-Paul Marbeau: The next speaker will talk about Gravity and Magnetics extensions to Epicentre. Why is that not on the Member Choice list? Answer: The incorporation of those extensions could be proposed as Work Category II project?
• Paul Maton: There is a plan to take the Gravity and Magnetics work that has been done and make it available to the industry. We will see if there is broad support to include the Gravity and Magnetics extensions in a future version of the POSC specifications.
• Jim Brannigan: The Member Choice work item list is not closed. The Gravity and Magnetics item could be added.
• Bertrand du Castel: Please take the input and feedback from this meeting into consideration. You are receiving valuable feedback now.
• Keith Steele: We want to understand that the Java DAE work item will not result in POSC developing a new DAE sample implementation. (Ed.: A new DAE sample implementation is not required.)
• <Conoco Attendee>: We have problems with well log service classes. How do you go about such reference data development project? Do you bring in people from SPWLA, etc.? Answer: Yes.
• Question: How long will the proposed work item on well log service classes take? Answer: It will be completed this year.
• David Archer: Significant work has been done already on this one.
• Keith Steele: You did have a plan to take specifications and create a real life demonstration to show the benefits. Has any progress been made on that? Is that considered as one of the Member Choice work items? Answer from David Archer. A work item for such a demonstration is being defined.
• Stewart Robinson: When you do make specification changes, will they go into Version 2.2 or into another version? Answer: That depends on the scope of the change. Changes could be delivered in a Version 2.x or in a new major release, Version 3.0.
• David Archer: The current baseline is Version 2.2.
• Dan Schenck: We are working to define a process that will allow us to deliver small, incremental releases.
• Steve Daum: We should ask POSC to describe the strategy for release definition at the next Member General Meeting.
• <Conoco Attendee>: Have POSC staff people been associated with each Member Choice work item so that specific feedback can be directed to the right person? Answer: Yes.
• Eric Hatleberg: We can add the names (as well as the work estimates) to the document on the Web site.

Project Report Sessions

E. FieldBank Project: Gravity and Magnetic Extensions to Epicentre; Kitty Hall (Ark Geophysics)

(PowerPoint Presentation)

Kitty described the FieldBank project, including the Gravity and Magnetics extensions to Epicentre, application software development, and data loading.

Questions and Answers:

• Rick Morneau: Are the gravity and magnetics survey traces (i.e. digits) stored in Epicentre? Answer: No. The data store is an index to the trace data. The data model extensions include support for storing the trace data in Epicentre. Question: Where is the entitlement data stored? Answer: It is stored in the data store. The data store at a site is password access-controlled. There is individual entitlement to the individual survey data.
• Gina Severeijns: We are doing similar work in other areas of data. We believe that the trace data should be stored in Epicentre. Answer: Phil Davies from PECC can explain the decision to store the trace data outside of the data store. This was a commercial issue. We can change the approach in the future.
• Nico de Rooij (as facilitator) ran a survey to record the views of the attendees as to recommend in which POSC Work Category the incorporation of the results of the FieldBank project would be proposed. Of 39 people voting, ten recommended a specification refinement work item in Work Category I, 27 recommended a funded development project in Work Category II. (Ed.: It will be up to Ark and/or other members to propose this work item. It would be appropriate to propose it for consideration as a Member Choice item in Work Category I. If that proposal fails to receive sufficient priority to be accepted, then a group of members can propose to fund the work as a Work Category II project.)

F. WIME Update; David Roper (Sysdrill)

(PowerPoint Presentation)

David explained that the objective of WIME is to validate the drilling and well operations topics of Epicentre. He said that the first focus of WIME is to define activity codes. This complements prior and concurrent work by POSC and POSC/CAESAR. The results are intended to be working codes and definitions to be stored in data stores and used in applications. David invited everyone to come to the WIME workshops, including the workshop on 19 June at Sysdrill in Aberdeen.
 

Questions and Answers:
 

• Nico de Rooij: Could the WIME work serve as the basis for modeling an audit trail in the geoscience disciplines? Answer: The WIME approach uses a very general activity model. Whatever you can put something on a GANTT chart, you can define use the same approach used in WIME. It should be possible to define planned activities and to relate actuals to planned.
• Bertrand du Castel: What is the mechanism to have the WIME results incorporated into Epicentre? Answer: We should ask POSC about that. We started with the intention of presenting the results of WIME as requirements to POSC. I believe that the incorporation of the WIME results should be proposed as Member Choice work items in Work Category I. The work may actually turn into a development project in Work Category II.
• Nico de Rooij (as facilitator) ran a survey to record the views of the attendees as to recommend in which POSC Work Category the incorporation of the results of the WIME project would be proposed. Of 19 people voting, twelve recommended a specification refinement work item in Work Category I, seven recommended a funded development project in Work Category II.

G. OpenSpirit Alliance Update; Nico de Rooij (Shell)

(PowerPoint Presentation)

Nico presented an updated on the status and plans of the OpenSpirit Alliance. He described the work that was done to prepare submissions to the POSC Interoperability and Business Objects RFT. He described the incremental releases of product, plans for demonstrations at the SEG Conference in September, and plans for commercial release of product by year-end.

H. RESCUE Alliance Update; Knut Kirkemo (Statoil)

(PowerPoint Presentation, requested)

Knut described the RESCUE project and concluded that the project has achieved a major achievement. RESCUE is a success. He said that the environment of cooperation between competing vendors has been really magnificent.

He described the series of RESCUE ILAB workshops. The recent ILAB5 compared various methods of storing large data sets and demonstrated that there is a large penalty of about three orders of magnitude when large data sets are stored as a large number of primitive entities.

Questions and Answers:

• John Sherman: It is probably possible to achieve good performance with a data model on a commercial basis equal to performance using flat files. I believe that to do so would lead to a solution different from the current Epicentre specification.
• David Roper: We didn't actually change the data model in the WIME project. We just took abstract concepts and projected in certain ways.
• Bertrand du Castel: There was a question from the members to the recent Board meeting about the future of the RESCUE project? Answer: The future of the RESCUE project is being discussed now. We must get together and set out objectives. I believe that it is important to keep the RESCUE initiative going. The next RESCUE business meeting has not been scheduled.

I. URGENT Update; Sheena Henderson (PDS)

(PowerPoint Presentation, requested)

Sheena described the status and results of the URGENT project. She raised a question from the recent URGENT Forum about whether investment in collaboration on the concept of Shared Earth Model (SEM) will go forward given the current low oil price. Statoil has a project for a corporate data repository using URGENT-defined work processes. AGIP took process change results for dissemination in the organization. Norsk Hydro have R&D projects for real-time updating and advanced reservoir modeling. All of these oil companies are changing their work processes.

The Forum concluded that 'as-is' processes have been modeled extensively. The methods and tools worked well. URGENT needs wider participation and verification. Moving forward with SEM will require multi-oil company collaboration. Synchronization of the URGENT work with POSC's work is desired.

The release strategy for the URGENT deliverable material has not yet been determined. Continued EC funding will be sought for more detailed work.

Questions and Answers:

• Stewart Robinson: Is the commonality among business process across the three oil companies due to the focus on Norway. Will the results extend to other oil companies? Answer: We had input from Italy as well as from reviewers from BP and Shell. We believe that the URGENT results are more generically applicable that Norway-Italy.

J. Shell/Oman Project Results; Gina Severeijns (Shell/PDO) and  Floris Versteeg (Oracle)

(PowerPoint Presentation, requested)

Gina showed a VRML demonstration of the Shell/PDO Oil Field Information work. Gina proceeded to describe the joint project between PDO and Oracle.

Floris described the technical project scope including the Oracle V8 features of interest. He described how the project performance-tuned the data store by defining indices for the POSC Epicentre V2.0 relational implementation. He talked about the concept of parallelism, which refers to dividing Oracle tables over multiple table space partitions. This technique provided good benefits. The project found that Oracle V8's treatment of objects were not mature enough to be used. The project used Java applets. The project used "Instead-of" triggers for data loading. The SDO features were used for spatial data, but efficiency problems were found. A major problem is that there is no support for 3D.

Floris concluded that Oracle V8 features must be evaluated for use individually. The project was successful. It had a good team and clear vision of the intended results. Shell/PDO also benefited from improved efficiency in the current Oracle V7 implementation by applied lessons learned in this project.

Questions and Answers:

• Gerard Huard: Did you address security? Answer: No. All data is available to everyone.
• David Roper: Were the index optimizations fed back to POSC? Answer: No.
• Gina Severeijns: We have published a detailed results document in draft. POSC were involved in the project and will receive the report.
• David Roper: Will the report be made public? Answer: I don't see why not.
• Michael Ring: A goal of this project was to achieve integration. Was that successful? Answer: This was not an objective of this project. That is the total goal of all of the data warehouse efforts in which all areas of data are fully integrated.
• Steve Trythall: Did you use reference types or collections in Oracle V8? How much performance benefit did you achieve from partitioning? Answer: In selecting from objects, collections were not stable in practice. We think we should wait until Oracle V8.1 and then try to use Oracle V8.1 with the DAE implementation. Yes, there were performance benefits from partitioning of about six times. The details are in the project results report.
ACTION ITEM J.1: Gina Severeijns will release the Shell/PDO - Oracle V8 project results report to POSC.

Project Proposal Sessions

K. Interoperability and Business Objects Request for Technology (RFT) Process Update; Alan Doniger (POSC)

(PowerPoint Presentation)

Alan described the activities that led to the current Request for Technology. He identified the submission received from combinations of eight POSC member organizations. He described the current effort of a member-staff team to evaluate the submissions. He anticipated publishing the results of the evaluation as soon as the end of July. There will be a public forum on the RFT evaluation results during the SEG Conference in September. Part of the evaluation results will be a work plan for the work required to complete the first set of POSC interoperability specifications.

Questions and Answers:

• John Sherman: Will the use of the architecture be separable from the use of the business objects so that alternative architectures may be substituted while still using the business objects? Answer: If I understand the question properly, then it will be the case that the business objects will depend on the existence of elements of the architecture For example, the intended use of a data object depends on the presence of units conversion services as defined in the architecture.
• John Sherman: There is a concern if the independence of the business objects is not preserved as defined in the ANSI/SPARC 3-tier architecture.
• Steve Trythall: ANSI/SPARC does not apply here. Both the services and objects are defined in the same architectural layer.
• John Sherman: We disagree.
• Alan: We should revisit this when the evaluation team report is available.
• (Ed. note: John's reference to the ANSI/SPARC data architecture indicates that we are not using the term architecture in same way. It appears as if John is referring to data architectures in the sense of data base organization paradigms, such as network, hierarchical, relational, and object-oriented. The POSC interoperability project is referring to application software architecture in the context of distributed object technology. In John's context business (data) objects are essentially external views of data. In POSC's context, business (data) objects are independent objects that can be implemented, configured, and addressed independently from application objects and data store objects. More dialog with John should take place on this point.)

L. POSC Project Proposal: Shared Earth Model; Dan Schenck (POSC)

(PowerPoint Presentation)

Dan reviewed the proposal for a Shared Earth Model project. He referred attendees to the proposal document which is available on the POSC Web Site.

M. Coring and Core Analysis Project Proposal; Bob Skopek (Petrophysical Applications International)

(PowerPoint Presentation, requested)

Bob reviewed the proposal for a Coring and Core Analysis Project. He referred attendees to the proposal document which is available on the POSC Web Site.

Questions and Answers:

• David Roper: Is the objective of the proposed project to make sure that the data model can handle all of the coring and core analysis information and activities and to populate best practices as standard data. Answer: Yes. The project is also intended to define standard formats for the exchange of core data.

N. Vendor Data Store Mappings Project Proposal; Steve Trythall (PrismTech)

(PowerPoint Presentation, requested)

Steve stood in for Geoff Thornley who was on his way to Australia. Steve explained that with more and more Epicentre implementations, people have the problem of exchanging information with the several vendor data stores in use. He said that there are a number of companies working on solutions of three types:

• point-to-point (e.g. Landmark's bridge product),
• applets (e.g. GeoShare, DAEX), and
• automation with formal mapping description (e.g. Migration Manager from PrismTech).
All of these approaches need formal description of the mappings. This project intends to define mapping descriptions for popular vendor data stores.

Questions and Answers:

• Stewart Robinson (to Bertrand du Castel): Have you mapped Finder to Epicentre? Answer from Bertrand du Castel: Yes. We did so with POSC, in fact. Answer: I believe that the production portion has been mapped and projected. I understand that all new domains of Finder are mapped together with POSC.
• Paul Maton: Does the proposed project intend to capture and use existing mappings? Answer: Yes.
• Dan Schenck: Many mappings already exist. Some of them may be in conflict with each other. This project provides is an opportunity to clarify existing mappings.
• David Roper: Why have two specific vendor data stores (GeoFrame/Finder and OpenWorks) been selected over all others? Answer: We are willing to have the project address any vendor data stores. The choice will depend on the preferences of the project participants.
• David Roper: This seems to be POSC Work Category III project that will deliver benefits to the project sponsors' direct commercial interest. Isn't this the same as what Oilfield Systems is doing? Answer: No. In fact, Oilfield Systems will benefit from the results of this project. The proposed project will not be implementation product.
• Gina Severeijns: Shell/PDO may benefit the most from the results of this proposed project. We need an industry-wide agreed mapping. Will the mapping be in both directions? It is important to get away from several mappings. I strongly suggest that we support this project.
• Nico de Rooij: What are the dimensions of this project? Answer by Dan Schenck: The first portion will use 2 to 4 people over six months. We thought of mapping more than one vendor data store at a time. The domain of choice is up to the participants.
• John Gillespie: Have you thought about the issue of version control given ongoing changes to vendor data store definitions?
• David Roper: It seems to me that POSC will burn money maintaining these mappings.
• Bertrand du Castel: We had better look at this proposal from a legal standpoint.
• David Archer: I don't see that it is a problem.
• Bertrand du Castel: We need a legal opinion on that.
• Bob Skopec: Are we encouraging vendors to stay with their data stores (as opposed to migrating to Epicentre) by doing this?
• John Sherman: The practical realities are that the vendor data store systems exist. They are deployed. It is useful for POSC to track the mappings. If there are multiple mappings, then there will be data collisions.
• David Roper: The only people that really understand OpenWorks are Landmark people.
• Stewart Robinson: Each installation is tailored. POSC should store mappings. We will do similar exercises. We want to share them.
• Nico de Rooij: How  will you do it? Answer: Individuals in POSC have been working on mapping methodology. Jim Theriot has a good paper on how to run a mapping project (on the GeoShare server).
• Gina Severeijns: If you take a limited approach and publish each one, that is good. If you try to map everything for all purposes, the project can't succeed.
• Dan Schenck: There is a page on our Web Site to publish mappings. The problem is that if we are not systematic, the quality of work can vary widely. We need some kind of consistent quality -- it may not be perfect. We may not keep them up to date all of the time.
• David Roper: I suggest that you concentrate on the formalization of the mapping language. Encourage others to do mappings. POSC should not get involved.
ACTION ITEM N.1: POSC will review project proposals, including the Vendor Data Store Mapping proposal, to ensure that there are no legal issues related to them.

Closing Sessions

W. Facilitator's Discretion;  Jenny Meader (POSC Consulting Ltd.)

It was suggested to have further discussion about Work Category I and Work Category II at a future meeting.

It was suggested to have further discussions about the relationship between POSC and non-POSC / POSC-related projects, e.g. those from OpenSpirit Alliance, SWPACT, etc.

X. Topics for Future Meetings

• Describe the strategy for release definition (from Session D)
• POSC's Work Category I and Work Category II, including actual and proposed work efforts with scopes and expectations, systematic updates on Work Category II projects, actual progress on marketing efforts, etc.
• Sessions should be included that build on the knowledge of the attendees, whether this is in plenary form or in break-out sessions.
• Sessions should be assured adequate time for questions
• The day's agenda should be reviewed in more substance at the beginning of the meeting to fully prepare the attendees for their role in the sessions.
• Possibly a session on version control
• The user expectations from a new release of the specifications
• Compliance plans and programs

Y. Action Items

ACTION ITEM C.1: POSC will publish the list of member organization by election category as part of the call for nominations issued in July.
ACTION ITEM C.2:. Member organizations with an interest in becoming POSC Certified Consultants will send to POSC a description of how the certification process should be defined. POSC will request this input from member organizations.
ACTION ITEM J.1: Gina Severeigns will release the Shell/PDO - Oracle V8 project results report to POSC.
ACTION ITEM N.1: POSC will review project proposals, including the Vendor Data Store Mapping proposal, to ensure that there are no legal issues related to them.