20090618 arb telcon minutes

Architecture Board

June 18, 2009

Attendance

Name	Present	With	Affiliation	E-mail address
Curry, Jane	Yes	ArB	Health Information Strategies	janecurry@healthinfostrategies.com
Grieve, Grahame	No	ArB	Kestral Computing	grahame@kestral.com.au
Julian, Tony	Yes	ArB	Mayo Clinic	ajulian@mayo.edu
Koehn, Marc	No	Guest	Gordon Point Informatics Ltd	Marc.Koehn@GPInformatics.com
Koisch, John	Yes	ArB	NCI	koisch_john@bah.com
Loyd, Patrick	Yes	ArB	Gordon point Informatics LTD.	patrick.loyd@gpinformatics.com
Lynch, Cecil	No	ArB	ontoreason LLC	clynch@ontoreason.com
Mead, Charlie	No	ArB	Booz Allen Hamilton	charlie.mead@booz.com
Nelson, Dale	No	Arb	II4SM	dale@zed-logic.com
Ocasio, Wendell	No	ArB	Agilex Technologies	wendell.ocasio@agilex.com
Parker, Ron	No	ArB	CA Infoway	rparker@eastlink.ca
Quinn, John	No	ArB	Health Level Seven, Inc.	jquinn@HL7.org
Shakir, Abdul-Malik	Yes	ArB	Shakir Consulting	ShakirConsulting@cs.com

Agenda

• Call to order
• Roll Call
• Approval of agenda
• Approval of Minutes of the June 11 Telcon June 11, 2009 minutes
• Out of cycle
• adjust timeslot to pick up more people?
• Update from EA Rollout Projects
  • Seminar Call
  • Breakout sessions
    • NCI
    • SD
    • CTS2
    •  ??
• Discussions of Platforms, Platform Definition Models, and PSMs
• Other business and planning for next call
• Adjournment.

Call to order

The meeting was called to order at 3:04pm U.S. Eastern with John Koisch as Chair, and Tony Julian as scribe.

Approval of Minutes of the June 11 Telcon June 11, 2009 minutes

 John will add topic headers to points.

Motion to approve minutes. Tony/Jane (3-0-0)

Out of cycle

John Koisch: put in formal request for out-of-cycle. Most would like to attend, and get funding for the August 18-23 harmonization/out of cycle.

John Koisch: 6-7 want to join. My sense is the HL7-HQ is looking for more output. We have gotten stuck on issues - and need to get deliverables out there. So the point is that the out-of-cycle is valuable, but HL7 is not sure if they want to do it.

John Koisch: Phone calls with John Quinn and TSC - John supports. John Koisch is pushing for it, but there is a question of funding.

adjust timeslot to pick up more people?

 John will send an e-mail to those who do not make call query best time/days for telcon

John Koisch: keep call to one hour

Jane Curry: time/day works for me

Patrick Loyd: time/day works for me

Tony Julian: Time/day works for me

We need to have other calls

John Koisch: We will keep this call an hour.

Update from EA Rollout Projects

John Koisch: SD is not opposed to work within the guidelines.

• Seminar Call

John Koisch: Have a seminar call work with Marc Koehn to present to the EA rollout projects a way to look at this, and evaluate the mapping of the artifacts to the SAEAF. This will turn out to be an implementation guide.

John Koisch: SD is interested with one or two projects. Patrick, I missed OO. There are quite a few projects, and I think a seminar would kick-start this.

John Koisch: Are there any problems with the approach?

Patrick Loyd: I thing it is a great idea.

Jane Curry: One persons question will spark another, and be more effective than a one-on-one.

John Koisch: The notions of the breakout sessions are to take this IG material to work with groups, at least offer our services. I dont have the bandwidth for all of this. Is there any interest in being a 'imbedded individual' for individual projects.

Patrick Loyd: yes

Jane Curry: Templates repository is a good fit. We are doing requirements analysis.

John Koisch: If we make it a little more concrete, and work with Marc's tutelage for management. The slides will re-constitute the slides we have. Hopefully this will be a helpful thing.

Jane Curry: What is the process for formalizing the EA projects, so Marc can track.

John Koisch: Mark is using the project stuff, asking for a charter, and rough timeline. I helped Russ Hamm with CTS2. This will put it on Marc's radar. The SD this morning could not write the charter - how do we nurture this, what do we do?

Jane Curry: Do you mean to do one per group, or do a seminar and invite mutiple groups?

John Koisch: Initially one call, here is how you do it. Then breakout sessions for each group. I have offered the initial call for SD. There are all these projects lost in the jungle, so we want to give them some light.

Jane Curry: Any idea when we will do the first one?

John Koisch: Before august out-of-cycle. Week after July 4, or the following. Gives groups a chance to make progress.

Jane Curry: Week of 6-10? SHould we start getting calendars organized?

John Koisch: I will have Marc coordinate. We will focus on groups that want to go down this road.

Jane Curry: limit on go-to-meeting.

John Koisch: NCI sentra has ability to do much more.

Patrick Loyd: Basic go-to-meeting only goes to 20 or so.

John Koisch: NCI has offered resources for this. If I can get this into Marc's head, I wanted to run by you guys before I pass it to him.

Jane Curry: I find july8 the most convenient time. Wednesday works well. July 15th I can do also, but PM mountain time.

John Koisch: You are in mountain time, right?

Jane Curry: Yes.

Tony Julian: Tuesday are bad for me.

Discussions of Platforms, Platform Definition Models, and PSMs

John Koisch: Should we break early. I tried to get a definitive definition from OMG. It is clear to me that in some ways that we are defining platforms - instead we are taking a broad set of assumptions to apply to a platform. V2 and V3 are platforms.

John Koisch: I have been working on the definition of platform as

"the specification of a common set of features of a technology in sufficient detail that a logical model may be consistently and traceably instantiated"  

Jane Curry: That means that the platform-independant level defines compliance to the business, information levels, and platform states how you will deliver it. You understand your platform, e.g. dotnet or java, using msMq or MQ series for java. The thing we have been missing is that you cannot create a platform-specific model withou knowing the platform.

Jane Curry: Right - we place the platform specification at that level, e.g. schemas are platform specific for xml.

John Koisch: using xml/java and xml/dotnet you do things differently. You build a platform on top of a platform, ultimately to the OS.

Jane Curry: Right. We need a picture of that.

John Koisch: I need to do a slide deck on the engineering viewpoint. I found it using the BF.

Jane Curry: Engineering viewpoint at a generic level - concrete example would be nice. They just dont know which slot contains existing artifacts.

Patrick Loyd: I think john left.

John Koisch: I dropped off. Sorry.

John Koisch: I will put concrete stuff to that, and lay it out.

Abdul-Malik Shakir: Some people think XML is platform independant.

John Koisch: I reached out to OMG - Richard Soey and John Segal - nobody can agree. I ran my definition by John. From OMG xml dotnet java J2EE.

Abdul-Malik Shakir: HMD is platform independant.

Jane Curry: it is one layer into the platform. I am thinking of the stack, trace an instance all the way to the bottom.

Abdul-Malik Shakir:

John Koisch: Is HL7 v2 and HL7 V3 platforms, yes from John Segal. Documents, messaging, and services are patterns that are applied against the platform specific model.

Abdul-Malik Shakir: I think that what we do in chapter 2 with decoding is platform.The idea that we have different information structures is platform independent. Sayin HL7 versin 2 is a platform does nto jive with me. The coding is.

John Koisch: Here is why it is: Any give message structure, you should be able to look at v2, and understand the encoding, fields, how they line up, an where to put each information type. Once you constitute it, and transmit it, the point is that you can create platform to platform. V2 is realized via technology on layer of platforms. The same with V3 -ceptualizing acts, messages. We are not building platforms with teh SAEAF, we are building patterns.

Abdul-Malik Shakir: using the label V2 or V3 as a platform. I can talk about it whetere I am going to send it xml or bar delimited. The platform independent view has nothing to do with the delivery.

John Koisch: So it is a platform.

Abdul-Malik Shakir: there is a platform independent view. V2 encoding rules are a platform.

John Koisch: HL7 v2 as a platform, or v3, they readh up the stack and impact the creation of the models - you build to suit. One of the distinctions you choose platform - messages, documents, services, and the ultimate platform specific model is the complete application of that.

Abdul-Malik Shakir: we agreeing on that there are layers. We are disagreeing on the dividing lines between levels.

John Koisch: Subsequent question - do we need to? Each layer has layers in it.

Abdul-Malik Shakir: when I look at V3, and RMIM, i put it in platform independent.

Jane Curry: Agree

John Koisch: But you are creating a message.

Jane Curry: not really.

Abdul-Malik Shakir: It is derived from the RIM, maybe it is platform.

John Koisch: RMIM for speciment management, describes the messages supporting an interactions. you have a topic, and an RMIM that suits a messaging model. An RMIM in general is a logical artifact.

John Koisch: Specimen management, you know that the platform is HL7 V3 messaging, this set the course for building the messages. SM RMIM could be modified to support services, with transformation, since some of the structures dont appear on the wire.

Abdul-Malik Shakir: Does that make it platform specific?

John Koisch: No,it is a logical structure built with a platform in mind.

Abdul-Malik Shakir: Designs are built with a paradigm in mind.

Jane Curry: Isnt paradigme the first platform decision,early in the design stage? Not a matter of specificity, but knowledge of the end-state. Design premises - we need to describe the constraints at the conceptual level, rather than saying we ignore it. you can create strictly platform-independant constructs, but how frequently, and how usefull to audience. Some have less impact or are more flexiably suite do one platvorm.

John Koisch: RMIM for Speciment management is a HL7V3 messaging model. You could create an object serialization, e.g. java beans. Point is that from the logical models you can create a platform specific, e.g. messaging/queue vs java-beans. Still conformant to the logical model.

John Koisch: V2 and V3 being platforms are a shortcut for describing the encoding rules. Against which we write platform specific models, which reach up the stack.

John Koisch: Send agenda items to me for next week.

John Koisch: AMS do you have time tomorrow to talk about the BF?

Other business and planning for next call

None

Adjournment

The meeting was adjourned at 4:00pm U.S. Eastern Tony 20:07, 18 June 2009 (UTC)