2009-05-21 Developers Conference Call

Date

21 May 2009

In Attendance

  • Mike Seaton

  • Ben Wolfe

  • Darius Jazayeri

  • Burke Mamlin

  • Paul Biondich

  • Justin Miranda

Agenda

  • Discuss agenda for meeting in Boston next Tuesday (Hamish can only be on the beginning of today's call, so we need to start with this, ideally with Hamish, Darius, Ben, Burke, and Paul on the call) (moved to afternoon because Hamish can't make it at 10am today)

  • Discuss 1.5 alpha release - Ben

    • Organized testing?

    • Discuss 1.5.x creation?

  • Testathon wrap-up - Justin/Ben

  • Discuss hibernate patch the Darius did to get to set concept ids on concept creation (not use auto generator) patch to hibernate jar

  • Discuss liquibase patch that Ben did to get imports working correctly patch to liquibase jar

  • Sync updates

Minutes

  • 1.5 alpha release?

    • Ben: The hibernate patch and update wizard have both been put into trunk. So 1.5 is ready

    • Mike: alpha should be released now that it is feature complete

    • Darius: we need a minor amount of testing

    • Mike: We don't need to spend a week testing an alpha...it only delays more things

    • Burke: We need to document this process of releasing an alpha

      • whats released when, where, etc. what is tested before/after cutting the 1.5.x branch, before/after 1.5 war release, etc

    • Paul: this discussion needs to focus on what is important...not all the other side issues of us not wanting to release buggy code, etc

    • Burke: the release manager for 1.5 can decide how to do things this time...just document it.

    • Darius: to-test: login, create patient, edit all patient attributes, create user, edit user password, create encounter, add/edit obs in it, void the whole encounter, create and edit a concept

      • Create database via initialization wizard

    • Make sure there is verbage in the release notes about not using this in production environments

  • Testathon

    • Justin: spent too much time on Saturday writing @shoulds. Next testathon these should be done ahead of time

    • Justin: Keelhaul, nribeka, glen_, ben, justin were writing tests on Sunday

    • Justin: we need to review the @shoulds and tests to make sure that they are doing what we expect, not just mimicking what the code currently does.

    • Justin: we don't have a convention for the @shoulds yet.

      • Burke: can you add what conventions onto the wiki

    • Ben: We can't expect community members to create @shoulds. once those were up there test creation was speedy

    • Justin: ran out of time to create the @shoulds ahead of time. next time we'll plan an @shouldathon before the testathon

  • Burke: TODO (Justin) we need to create a ticket to create a ticket for creating a Selenium framework in openmrs (or tell us why its not feasible)

  • Burke: TODO (Justin) modify the unnamed link wiki page to add the conventions

  • Hibernate patch that Darius did

    • The identify generator does a .equals comparison instead of a .isAssignableFrom so identity generators are not able to be extended

    • Darius: the hibenrate dev list never got back to him about the "bug" and so he couldn't create a ticket

    • TODO: (Darius) create a ticket in hibernate for this patch

  • liquibase patch that Ben did

    • he hasn't applied all of the first patch

    • TODO (Ben) link to the liquibase ticket for the patch

    • TODO (Ben) make tickets for the other changes

    • Burke: patching other libraries code needs to be a group decision. if the group decides to put off development or to move ahead with the patch

  • sync updates

    • getting 1.5 out the door (because it has uuids on it) is part of the process

    • Darius: sync-by-usb-stick has not been tested

    • Burke: we could potentially test sync-by-usb-stick in kenya fairly easily because similar tests were done for remoteformentry

    • Ben: tickets were created for the two tasks that need to happen in trunk (required modules and allow modules to add an interceptor)

    • Darius: maros is studying for the MCAT, so he hasn't been working much on sync module conversion