Notes from Joint XD* Strategy Meeting
Date: Friday, October 8, 2010
Time: 1:30pm-2:15pm
Attendees: George Cole, Ron Cordell, Sri Koka, Vince Lewis, Arien Malec, Vassil Peytchev, Jas Singh


Vince Lewis
  • Informed group that he is conducting clean-up as well as re-factoring
    • Should be done by the middle of next week
Arien Malec
  • Recognized that Beau is working on the object model
  • Asked if Beau Grantham had checked in the resolver
Vince Lewis
  • Responded that Beau Grantham has prepared a sub-resolver and interface
Arien Malec
  • Indicated that is a good way to do it thus far
Vince Lewis
  • Suggested they iron out the business logic to the resolver
Arien Malec
  • Indicated it would be based on a configuration where you always know the domain
    • With respect to sending the XDR
    • Otherwise you send the SMTP
Vince Lewis
  • Asked about the idea that resolver is supposed to determine the endpoint
    • Not just determining addresses
Arien Malec
  • Expressed agreement with Vince Lewis' understanding
  • Spoke of a hand-written implementation
    • Suggested looking at the MedAllies implementation
Vince Lewis
  • Agreed with Arien Malec's perspective
Arien Malec
  • Indicated that over time they can swap it out to UDDI without changing anything
Vince Lewis
  • Expressed goal to not create something too cumbersome
    • If he is sending SMTP, then he needs to send to that endpoint
    • However, if kept simple then there should be no problems
Arien Malec
  • Transitioned to the CSharp side
Vassil Peytchev
  • Clarified that he has been out of the loop for a little while now
  • Indicated he had been pulling together an object model for the XD* metadata
    • Generating XD metadata when there is a CDA payload
      • Plans to merge this
  • Recognized that Arien Malec has created a similarly functional XD* object model
    • Generated without SOAP
    • Once he merges his item into the repository
  • Determined to hold a call with Arien Malec to determine which direction to go
  • Stated that they will be pulling from the codeplex
    • Not sure if anyone worked on that since the Code-a-thon
Arien Malec
  • Responded that no one has worked on it since Ron Cordell put that in
  • Clarified that George Cole will be provided some code in smaller packages
    • Recognized there may be two competing object models
George Cole
  • Stated that a single project was checked-in which should address some of the XDS
    • Registry error
    • Registry error list
Vassil Peytchev
  • Asked if this was an XD or XDS file (unresolved)
Arien Malec
  • Clarified that he does not have a sense of ownership
    • At the end of the day, if they can get an implementation up - that would be preferrable
      • Document entries - registry package mission sets
  • Suggested they decide between Vassil Peytchev's and Arien Malec's approach
    • Need to determine the correct path to go down before hooking up to the web configuration
    • Not sure that they need all the XD dependencies
Vassil Peytchev
  • Agreed with Arien Malec that not all the dependencies are important
    • If they can get the web service to generate the object graph through parsing
      • Then they can move forward
Sri Koka
  • Indicated that he was not able to do as much
  • Stated that he had looked at Arien Malec's code
    • Asked Arien Malec if he used LINQ
Arien Malec
  • Responded yes
Sri Koka
  • Indicated that he does not want to do this from the front
Arien Malec
  • Recognized that they have a couple of decisions to make
  • Asked Vassil Peytchev if he was free for Monday
Vassil Peytchev
  • Responded that he would be free after 11:00am EDT
Arien Malec
  • Decided to hold a CSharp XD* architecture meeting
    • Invited Vassil Peytchev, George Cole, Sri Koka and Jas Singh
  • Indicated that by Monday his object model should be able to "round-trip"
    • Is XDR code more compatible or is the parsing option preferred?