20090903 arb telcon minutes
Architecture Board
September 3, 2009
| Name | Present | With | Affiliation | E-mail address |
| Thompson, Cliff | Yes | ? | ontosolutions | cliff@ontosolutions.com |
| Curry, Jane (Goverance Girl) | Yes | ArB | Health Information Strategies | janecurry@healthinfostrategies.com |
| Grieve, Grahame | ? | ArB | Kestral Computing | grahame@kestral.com.au |
| Julian, Tony | Yes | ArB | Mayo Clinic | ajulian@mayo.edu |
| Koisch, John | Yes | ArB | NCI | koisch_john@bah.com |
| Kreisler,Austin | Yes | OO | SAIC - Science Applications International Corp | SAIC - Science Applications International Corp |
| Loyd, Patrick | Yes | ARB | Gordon point Informatics LTD. | patrick.loyd@gpinformatics.com |
| Lynch, Cecil | ? | ArB | ontoreason LLC | clynch@ontoreason.com |
| Mead, Charlie | Yes | ArB | Booz Allen Hamilton | charlie.mead@booz.com |
| Nelson, Dale | Yes | Arb | II4SM | dale@zed-logic.com |
| Ocasio, Wendell | ? | ArB | Agilex Technologies | wendell.ocasio@agilex.com |
| Parker, Ron | ? | ArB | CA Infoway | rparker@eastlink.ca |
| Quinn, John | ? | ArB | Health Level Seven, Inc. | jquinn@HL7.org |
| Robertson, Scott | Yes | OO | Kaiser Permanente | scott.m.robertson@kp.org |
| Shakir, Abdul-Malik | Yes | ArB | Shakir Consulting | ShakirConsulting@cs.com |
| Smith, Karen | ? | ArB | Technical Writer | karen@smithspot.com |
Call to order
The meeting was called to order at
Agenda approval
MMS to approve the agend below <Patrick/Tony> Vote:(7-0-0)
- Call to order
- Approval of agenda
- Discussion of OO
- There were concerns that OO was not satisified with agenda for q4 wednesday.
- Discussion of OO
Patrick Loyd: The agenda for Q4 is OK. OO decided that they would send one representative, since John is coming to Q3. OO will continue to work on dynamic model, if OK.
John Koisch: Charlie, do you see any concern.
Charlie Mead: Please repeat.
Austin Kriesler: OO is meeting q3/4 on wednesday for BH. Q4 they will send a representative.
John Koisch: Q4 is the big rollout meeting.
Charlie Mead: I met with Marc, and we hammered out the Q4 meeting. Marc would like each project would give a 5 minute talkd about what they are doing, what they know or dont know. The OO plans should not be a problem. I assume OO is interested in SAEAF from a messaging perspective.
Austin Kriesler: Model consistency across the board.
Charlie Mead: There is no concern with parallel meetings.
Patrick Loyd: OO agreed Austin or Patrick would be the best selection.
Charlie Mead: Good
John Koisch: Potential confilict - if OO is doing an EA rollout project, you will have to do the update.
Austin Kriesler: Will you be able to join OO in Q4?
Charlie Mead: You should be doing that John.
John Koisch: Can you do the 5 minute on SMS?
Charlie Mead: Yes
Austin Kriesler: We are good to go then.
John Koisch: Any issues?
Austin Kriesler: Yes, but not to discuss here.
- Discussion of SOA WG
John Koisch: In Traverse city we had discussions around governance. Charlie update ArB with TSC discussions. Groups want SOA to mentor them to build services. Emiled them to coordinate with ArB facilitator. Charlie update HSSP discussion with TSC.
Charlie Mead: I do think that the TSC is more aware of potential issues: Woody is worried that SOA workgroup are saying things about architecture of HL7, but should come through ArB. Last week on TSC had formal discussion about TSC step up and motivate./require things to happen. Which WG's will have to work with SAEAF. HSSP issue is one of the general issue of services, publishing EHR services standard. I asked Galen and have received no reply. Tsc will be requiring the two groups to participate, and to start the discussion about HSSP. The whole issue of governance will be dealt with in Atlanta. JK gave a good description of what is going on. It will be addressed.
John Koisch: There are action items for the TSC in Atlanta.
Austin Kriesler: There was a negative reaction in StrucDOc about being required to participate in SAEAF.
Charlie Mead: what will happen i think is that the TSC will not allow CDA-R3 to develop outside of SAEAF. There is a disagreement among SD co-chairs.
Austin Kriesler: The whole concept of anyone having governance of CDA is not well met.
Charlie Mead: It is needed for a EA.
Austin Kriesler: Just wanted known that there is resistance.
Charlie Mead: It is a TSC responsibility.
Austin Kriesler: Pushback that the TSC should not doing that.
Charlie Mead: The S in TSC stands for Steering.
John Koisch: There is the same issue with CTS2.
- Discussion of CTS 2
John Koisch: A lot of discussion in CTS2, common terminology service-2, balloted as DSTU,and passed, and is viable for OMG to certify for HSSP. Lloyd and Grahame have valid documented concerns about the DSTU and HSSP processes. There is an informal resolution from VOCAB concering Grahame and Lloyd concerns. This is aproblem that the process is flawed thatyou cannot fromalize good citezenship.The solution I proposed to CTS2 including Vocab WG
- OMG healthcare specs must be SAEAF compliant
- any project teams moving stuff will need review by HL7. This may require full ballot, or formal peer review. It should be run/facilitated by ArB. There need to be a revisit of the MOU between OMG and HL7. Lloyd was pleased by the solution - he wants to make sure things are going forward. This is the common thing with multiple project teams. It comes down to the whole compliance/conformance argument. I would like to hear an endorsement of that.
Austin Kriesler: Has implications - constrains the board.
Charlie Mead: What constraints?
Austin Kriesler: Constraints on the MOU's to follow SAEAF.
Charlie Mead: There is discussion in the governance about INTRA and INTER governance. John is seeing it extended beyond HL7, but HL7 should not micro-manage it. It is a framework, and components derived should use the frameworks. JQ was tasked to touch base with Richard Soley goving hinm a summary of the important things. Richard wrote back a committment to do that. OMG is tighter ran that HL7. Richard's communication is a good indication. AMS: The MOU with OMG does not go to the level of detail where this would make a difference: Most detail joint projects, membership agreements, etc.
Charlie Mead: There is a General MOU with OMG, as well as the HSSP process.
John Koisch: Push back: There is a reason to revisit the MOU. HL7 wants to say that if the other specifications are going to apply to healthcare, SAEAF will have to be adhered. E.G. CTS2 represents industry - if you dont formalize the relationship, you wont get compliance. AMS: Im not sure it is in the MOU. IF the Platform specific is developed in OMG, is it their product?
John Koisch: It is a shared model. The HL7 Conceptual model is HL7, OMG's product is OMGs, to bring back to HL7 as a normative standard. AMS: Bringing it back, and balloting, would come out with a normative HL7 product. HL7 develops, geta approved by ISO, then becomes and ISO product. Will the OMG then become HL7 product?
John Koisch: no one knows the review process. AMS: OMG may have constituents beyond healthcare, and not be constrained by healthcare/HL7.
John Koisch: What happens beyond OMG, e.g. CTS2, see a model beyond healtcare, and OMG does not agree with concept domain. HL7 needs the domain. There may then be multiple domains to a model. AMS: Which may come back to HL7 as a normative spec.
Jane Curry: Concept domains must be constrained to a value set. This alows compliance to a normative spec. AMS: We may get into similar with X12, NCPDP, not just OMG.
Charlie Mead: You dont know someone, then you see them on billboards - it will soon be 'Jane Jane the governance girl'
Jane Curry: It will also be the ArB's role to be part of the appeal process. More in the industry. Charlie had an IBM whitepaper about TOGAF. We need to understand what they are doing. They have expanded dramiticly the last six months.
John Koisch: Motion to formalize the two stages? Whatever is appropriate for each organization, that the MOU contain SAEAF language, and an ArB IPR process, maybe supplementary ballot review. Proposal on this call, or Atlanta.
Charlie Mead: Worthwhile discussion. TSC will discuss in two weeks.
Jane Curry: Agenda for Saturday is the other place we should come forward. I will be at TSC on saturday.
John Koisch: ArB should have draft thought.
Charlie Mead: TSC discussion on 9/14 should come first, then discuss and take to the TSC on 9/19.
John Koisch: ArB vote on the 17th
17 agenDA - discuss TSC discussion.
John Koisch: Not today, wait for the TSC to discuss.
- Primitives versus Composites
John Koisch: Jane set frame work for primitives/composities.
Jane Curry: We need closure on the GRID. While you break down the conformance statements, you can sort them into viewpoints, but some cross viewpoints. When we are dealing with composite artifacts, how do you understand which are primitive to reconcile for consistancy across viewpoints.
Charlie Mead: The thing that is important for a framework that the distinction between composites and primitives are important in governance. We cannot name the composites up front - as the Alpha projects execute and we review their artifacts, we will not know what we are talking about. Ther will be a lot of abstraction up front, then less.
Jane Curry: That means that our homework on the GRID can be closed saying these are premature.
Charlie Mead: We should fill in the grid for composites for the Alpha projects. IMO If we can identify primitives, we should, but focus on composites.
Jane Curry: I am struggling with the artifacts that live in the business viewpoint are the scope for implementation.
Charlie Mead: Principle point of focus.
Jane Curry: when it belongs to which group.
Charlie Mead: Which is why TSC wants CDA R3 to develop under SAEAF. There will be document primitives.
Charlie Mead: NCI has developed primitives at the computational viewpoint.
John Koisch: I agree. I think that the BF pints the tht fact that the V3 framework is working with the wrong primitives. The BF tries to draw the boxes differently. The question is less primitives/composite, instead think of analysis and systhesis.
Charlie Mead: I agree. The analysis will show composites/primitives.
John Koisch: One persons composite is another primitives.
Abdul-Malik Shakir: It sounds like primitives should fit in one cell - if spans, it is not primitives.
Jane Curry: Not quite true - there are connector primitives that connect concepts between boxes.
Charlie Mead: What you just hit on is the difference with RM-ODP and Zackman. The former on Composites, the latter on primitives. Relationships form another column.
John Koisch: they influence each other to the degree that they define an interaction to responsibilities. They go back and forth.
Charlie Mead: One of our principle areas of focus - filling in the grid is important for the alpha. We will care about primitives/composites later.
Charlie Mead: Can we move on?
John Koisch: We have hit our 4 topics.
Motion to adjourn at 3:54pm Tony 20:10, 3 September 2009 (UTC)
