RIMBAA 201010 Cambridge Minutes

• Focus this WGM: If one has an RIM based object store, what services should one define? And how does one query it? - RIMBAA will be duscussing ideas and experiences.
• Short URL: http://hl7.me/ro

Monday Q4 (15:30-17:00) in Paul Revere A

1. Approval of agenda for the week (5 minutes)
2. Administrative agenda items (max 30 minutes)
   1. Agenda Review/Additions/Changes
   2. Approval of the Minutes of the previous meeting in Rome: (max 5 minutes)
   3. Announcements
      • Canada Health Infoway have a fall partnership conference on November 15-17. This meeting contains presentations on RIMBAA topics, as well as a v3 implementation tutorial. Lloyd McKenzie, Lorraine Constable and Andy Stechishin will be in attendance, and have indicated they'll provide feedback to RIMBAA.
   4. Review WorkGroup Health - see [1].
   5. Report from the Rome WGM (Rene)
      • New concepts were presented by Michael and Ewout - they have updates during this meeting.
      • Discussion of javaSIG2010 - needs additional volunteers to make it available as open source.
      • Discussion of a Data Types library, and the criteria that it should preferrably adhere to.
   6. Update RIMBAA three year plan, whitepapers/due dates
   7. Update of the whitepaper project, result to PMO
   8. Planning of next meeting
      • London UK (November 2010)
      • Discussion on organizing a RIMBAA meeting in the US between the January and May WGMs.
3. Update on the joint SOA/RIMBAA Services for RIMBAA project (Ann Wrightson, SOA co-chair).
4. The use of data type level Schematron (Alexander Henket, max 20 minutes)
   • Nictiz, the Dutch NHIN profiler/enforcer, uses a set of Schematron files to validate the constraints in as documented in the abstract Data Types R1 specification. The constraints are actually those that apply to the Dutch data type flavor, but the mechanism is generally applicable.
     • Alexander will provide an overview of their approach, and will speak to its successes as well as its problems.
     • Their approach supports the full data type specialization hierarchy: the schematron for IVL_TS extends the schematron for IVL which extends the schematron for SXCM which extends the schematron for QTY which in turn extends the schematron for ANY.
     • They are in the process of creating CMET/Message type specific schematron files that use the data type schematron and extend them by business-rule validations. This approach has some limitations: schematron doesn't allow pattern-nesting, and creating an MT-schema to MT-schematron transform is a tricky process because of recursion.
     • Note that Nictiz is a Real World Implementation® using v3 artefacts derived from a mixture of v3 publications, and as such there are no MIF specifications to drive the generation of (e.g. Java) classes nor Schematron. The validation tooling is entirely XML-based.
5. Template based CRUD services (Michael van der Zel)
   • "Simple Store", Template based CRUD Services (Create/Read/Update/Delete, basic services) approach, one Service Decomposition in RIMBAA Applications pattern

Tuesday Q6 (19:00-21:00) in Paul Revere A

1. Focus: this quarter focuses on SOA v3 implementation experiences
   • NOTE: The meeting will have remote participant(s) (one person has signed up). Skype will be used BOTH for audio sharing as well as desktop sharing.
2. Implementing hData (Max 30 minutes, Gerald Beuchelt, Mitre)
   • Presentation on the topic of "mapping hData to in-memory RIM-based object structures" (the CS-CO or CS-RO cell transition in the technology matrix). The presentation will focus on the implementation of hData - and not on an explanation of hData itself.
   • hData is (using HL7 speak) a new XML ITS. See HData for an introduction, and www.projecthdata.org/ for detailed specifications, schema and FAQ.
   • How does this ITS compare to other ITSs when it comes to issues like code generation, code re-use and content validation?
3. Use of CMET-like CRUD entity services in the Nijmegen University Hospital (the Netherlands). (Max 30 minutes, Ewout Kramer, Furore)
   • Ewout: I construct them by comparing domains and then figuring out how to split the domains into manageable entities. These entities are small, and I also use a kind of reference indicator so one entity can refer to another entity. Large interactions are thus split-up (decomposed) into smaller, independent crud entities that are stored separately and can only reference each other. This way I try to solve the Object nets versus object trees storage issues. It also makes the store much more approachable and documentable for developers new to v3, I hope. I am storing each entity as a unit in its XML form in an eXist XML database. Splitting transactions and reconstructing them is done using XQuery.
4. Overview of the v3 SOA RIMBAA implementation at NCI (max 60 minutes, John Koisch/Jean Duteau)
   • NCI has embarked on the creation of a v3 SOA (SAIF based) based v3 implementation. The resulting code will be published as open source - it is funded under the US ARRA funding.
   • The solution will both support a OLTP database as well as an OLAP (object graph; for clinical/DSS queries). The OLAP database is based on the RIM. The services use "CMET-like" v3 models.

Thursday Q4 (15:30-17:00) in Paul Revere A

• Focus: this quarter focuses on open RIMBAA issues, notably those related to the persistent storage of RIM-based models.

1. Persistence models, LDM/PDM
   • See modules S760, S780 and S785 of the RIMBAA best practices training available at www.ringholm.de/download/HL7v3_implementation.zip
   • See these issues on the wiki: Implementation aspects of RIM based database models, ORM best practices.
2. Querying of subsets of persisted object graphs
   • See Object nets and object trees, Safe querying of a RIM-based data model, and Extracting data from a RIM-based object store.

Friday Q1 (09:00-10:30) in Paul Revere A

• Focus: Wrap-up, methodology issues surfaced by RIMBAA implementations