2007-07-06 Sync Conference Call

Two immediate PIH scenarios that must be addressed asap:

  • Lesotho
    • no patient overlap
    • form import only @ parent
  • Rwanda
  • What if we had a 'dumb' sync. It would have the following steps:
  1. Please 'sync' - initialize sync
  2. Figure out the last sync timestamp
  3. Find all new/change records since last sync
  4. Pkg the changeset
  5. Child: Send to parent
  6. parent: receive data from child
  • How do people feel about moving forward: consider these options:
    • a) no history at all
    • b) sync
    • c) do it all right from the start: history from get go, cleaning up the data model

Here is how people feel about it:

  • Christian: do c) to begin with; it will be in the end to do it faster
  • Darius: start now, and add history later – it will improve the code, should be possible but didn't prove it

    Julie: can't see how we can do it without: we*will run into the conflicts and there will be no way of resolving them

  • Justin: in for c)
  • Anders: prefers c)
  • Maros: can be done in step-wise fashion do b) and work towards c) concurrently
  • How do we move forward?
    • Darius: it is up to us (Darius and Maros) to 'prove' to rest that it will work
    • Maros TODO: write down the details of the sync steps and present to groups Monday @ 3PM EST