20100117 arb minutes
Architecture Review Board January 14, 2010
|Curry, Jane||Yes||ArB||Health Information Strategiesfirstname.lastname@example.org|
|Grieve, Grahame||Yes||ArB||Kestral Computingemail@example.com|
|Julian, Tony||Yes||ArB||Mayo Clinicfirstname.lastname@example.org|
|Koehn, Marc||Yes||Guest||Gordon point Informatics LTD.||email@example.com|
|Koisch, John||Yes||ArB||Guidewire Architecturefirstname.lastname@example.org|
|Loyd, Patrick||Yes||ARB||Gordon point Informatics LTD.||email@example.com|
|Lynch, Cecil||No||ArB||ontoreason LLCfirstname.lastname@example.org|
|Ocasio, Wendell||Yes||ArB||Agilex Technologiesemail@example.com|
|Parker, Ron||Yes||ArB||CA Infowayfirstname.lastname@example.org|
|Platt, Joe||No||Guest||Ekagra Software Technologiesemail@example.com|
|Quinn, John||No||ArB||Health Level Seven, Inc.||jquinn@HL7.org|
|Robertson, Scott||No||Guest||Kaiser Permanentefirstname.lastname@example.org|
|Shakir, Abdul-Malik||No||ArB||Shakir Consulting||ShakirConsulting@cs.com|
|Smith, Karen||No||Guest||Technical Editoremail@example.com|
|Thompson, Cliff||No||Guest||OntoSolutions LLCfirstname.lastname@example.org|
Call to order at 9:00am U. S. Mountain Standard
- Sunday Q1
- Call to order
- Approval of Agenda for Sunday
- Approval of Agenda for WGM
- Approval of Minutes
- Review of Decision Making Practices
- TSC update (Ron Parker)
Charlie Mead: We (arb)submitted and voted on membership criteria. TSC made non-substantive changes, which was published. TSC voted on and passed. Changes were TSC control over membership: preponderance of membership and perception of competance.
Tony sent an e-mail to Lynn to obtain.
Charlie Mead: TSC was active and productive for ARB and SAEAF. Discussion over the course of day about governance, and how it will happen. Issues about timing of SAEAF. ArB is not doing a good job of disseminating the SAEAF. To adopt takes top-down, and to implement takes governance. We need to do Architectural Harmonization.
Charlie Mead: NCI believes that SAEAF is intellectual property of HL7. Multiple organizations will have implementation guides. HL7 needs to worry about it's knowledge, and the HDF. NCI will support the tools to manage, and share the tooling to manage the SAEAF-derived semantic comtent, which they will share with the larger community.
Charlie Mead: We need governance: What we need to do is to go back to TSC with concrete notion of a architectural harmomnization would look like. How do we capture, and manage it. Discussion was held about the aspects of SAEAF and TOGAF.
Wendell Ocasio: There are a significant numbr of items in the stack that are addressed in both.
Charlie Mead: There is no real need to even do a formal alignment. The excercise is to make sure your framework is covered by both.
John Koisch: What does TOGAF say about conformance? SAEAF is an instance of TOGAF that focuses on interoperability - for a given community.
Andy Bond: SAEAF is not an instance of TOGAF. We need to enable the instances to exist, and interoperate.
Ron Parker: Challanging for us to maintain SAEAF as an alternative or form of enterprise architecture - it is risky. Produce aretifacts that will export to HL7, to be used to build architecture. (Insert rons drawing 1)
Charlie Mead: Focus on interoperability. TSC discussed the name again.
Ron Parker: Causes me angst - having to change the name to convey the meaning. People are desparate for coherent strategy for interoperability. We are grossly over-selling this - we need to tone down the rhetoric.
Charlie Mead: Closing the gap between rhetoric and reality.
Ron Parker: We need to discuss communication strategy: Enterprises do not need to build saeaf - they need to rely on a robust SAEAF expression of working interoperability(WI) to achieve it.
Andy Bond: Do not use SAEAF and enterprise architecture in the same breath.
Ann Wrightson Whitepaper on SAEAF vs TOGAF would be useful. RB: My experience with TOGAF will not change SAEAF.
Jane Curry: SAEAF is a filter. RB: Zachman is about business architecture. TOGAF will not solve the problems SAEAF addresses - and vice-versa. SAEAF is to deliver interoperablity.
Ann Wrightson COnclustion that it offers a toolkit, that will produce something like SAEAF.
Ron Parker: Premise is that I have a project, and I need to use achieve WI. Not Radically different, we are discussing the WI. There are huge implications of the understanding - business needs to be able to determine where it fits in the SAEAF.
Ron Parker: How does HL7 show derivation from the SAEAF? Is XYZ in the requiremenst? Where in the SPec?
John Koisch: At NCI have undertaken to understand HITSP interoperability spec( not core). A lot of alignment to the Behavioral Framework. Heart of it is traceability. Must be generally expressed in terms of EHR-FM, and how do you express in services? Work is maturing. Will share with ArB when ready.
Charlie Mead: Goal is to get SAEAF into the federal health mindset.
- Behavioral Framework (John Koisch)
John Koisch: document is not ready.
Jane Curry: How mature is it?
Charlie Mead: Not a question of content shifting, concern over unclear specifications.
Wendell Ocasio: What level of specificity is needed?
John Koisch: What are you trying to establish with a BF? HL7 has been focused on information exchange. My question to Charlie is, given people misunderstanding, what are we trying to get out of doing the BF?
Charlie Mead: People will say it is too general, insufficient. Woody is vehement about why it is not being used today.
Patrick Loyd: We have methodology at a 10,000 foot view, but not at the lower levels.
John Koisch: Woody wants to see the schema to go in the mif.
Ann Wrightson People in the UK want a short, lucid presentation of what is in it. The simple version should have been written first.
Patrick Loyd: we need to define a slice of artifacts, going to several groups, and one group must coordinate. We have three facilitators on OO. We agree that the whole list is invited to discuss.
Ann Wrightson When will we see the ballot?
John Koisch: Today, the BF was comissioned for the ArB to do. I was trying to get vote from ArB - content has not moved in 9 months. The BF is suitable for having the same discussion, of how enterprise a and B support busines process. Does not mean it supports a MIF schema. The BF is not ballotable - it is generic - and is like looking at TOGAF.
Ron Parker: NOt a normative artifact.
Patrick Loyd: Need to ballot the framework in some form. At the same time we will need to ballot the architecture. MIF: I with ron and lloyds help defined requirements for model.
John Koisch: We also have AMS models.
Patrick Loyd: I modeled the requirements.
John Koisch: what the BF is right now, vote on whethere it expreses how HL7 needs to define conformance on interoperability. SAEAF - ITS expression, and MIF schema - should be compliant. We have not covered how we will ballot a service?
Patrick Loyd: What are the pieces?
John Koisch: We have an ontology.
Ron Parker: Question - What constitutes sufficient evidence that we can declare due diligence. The content is stable - the problem is how to land the spec, with how it can be achieved. There are others ready to consume the BF without HL7 spec. Question is this:(BF is not in DITA yet).
Ron Parker: What constitutes due diligence?
Grahame Grieve: I dont think you will like my answer. We need a normative ballot using it. Until we have done that, we may have to re-visit the framework.
Ron Parker: It worries me. In Canada we had gaps between the spec and the anticipation. We need to get it out.
Grahame Grieve: Clearly the process requires a process. You cant get there until you do it.
John Koisch: What do we ballot?
Marc Koehn: Once you have done that, I suggest that we are worried about reaction, but what I am hearing is that we have a clear idea which we have not communicated. Today it is at this stage - and has three more steps to go through. Every one of the three chapters will go through say three steps to validate, and we have not written it down.
Jane Curry: We have committed to go DSTU before we go normative.
Charlie Mead: We are conflating two issues - communication, and the notion of needing ballot. Having been 4 months of ECCF documentation, with edits, it is not perfect. I am 150% in favor of getting it out. The first round of comments, and did edits. Why would we vote on the BF?
Patrick Loyd: There is a high level view, and a down to earth view, what do we vote on?
Ron Parker: Vote on language and notion on use, with the evolution explained.
John Koisch: We vote on the document to agree on how it is represented. It will move a bit. Vote as a governance step - that the path of a framework is the path of a framework.
Ron Parker: Will not commit unless they have sense of stability. What do you (arb) members need to verify the stability.
John Koisch: BF is the part of SAEAF that communcates the computational framework. Maybe we dont need to vote on it.
Ron Parker: When can we put it in the DITA. THis meeting we will notify publishing of the content.
Charlie Mead: You are conflating what we decide and what publishing decides. Karen does not want to put something in DITA if there are substantive changes. We can agree, but if there are 50 unanswered question, Karen will not put it in.
Jane Curry: DITA is focused on topics - if we are still re-organizing, then putting it in DITA will drive you nuts. Once the topics are mapped, we can wordsmith the items under the topics.
John Koisch: IMHO the questions out there are not valid - we need a release so we can go further.
Ron Parker: Are there question that the ArB has?
Jane Curry: There are comments we need to review. The process has been to go back to the questioners to see if the re-wording is acceptable. One reviewer is seeing inconsistancies. I cant tell if the questions have been addressed. I would ask Charlie and John, can we get a version with all of the questions resolved? The structure re-work is a huge effort.
Charlie Mead: John has a few figures to fix, and the comments from the 5 reviewers have been addressed.
Jane Curry: Need EA to save as svg. Can use EMF. AMS has a set of models, that needs to be expressed in EMF.
John Koisch: Model has been exported as HTML.
Wendell Ocasio: DITA needs vector graphics.
John Koisch: Load up the tools, so you can get at the EAP file.
Jane Curry: Karen does not have tools.
John Koisch: I will do it. If there is a image format issue, the consumer should solve it. You want people to work with the elemental things.
Wendell Ocasio: I dont need to have visio to look at the RIM models, what is the diffence.
Ron Parker: What is the source of truth. If we have to regen for each change, John would rather regen the core files, and the consumer will change to the view they need. This is a publishing excercise.
John Koisch: I will send it to Karen.
Jane Curry: As raw graphics, so they can be added as appropriate.
Ron Parker: We need a business process on methodology on content change and expression.
John Koisch: You understand my concern.
Jane Curry: We just need to understand the toolkit. We need to make changes in diagrams.
Ron Parker: Need to understand fit between SAEAF and TOGAF.
The meeting was adjourned at 11:00am U.S. MST