Retrospective Notes 2018.02.14

Wiki > Project overview > Notes from Retrospectives > Retrospective Notes 2018.02.14

Sprint_2018_01_18

Date of Sprint Retrospective: 14-02-2018

1. Testing API Changes

   • we made a change to genie_python (changed enumerated types) but did not create unit tests to test the changes. This came back to bite us.
   • we had a similar issue with #2942 - changes to DB records had impacts elsewhere, which caught us out.
   • Lesson: make sure there are tests that test your changes.

2. SM300 Motor

   • We estimated 8 points for this (thinking it was a big ticket). In reality, it required much more effort than 8 parts.
   • Creating a motor_record was more complex than we expected.
   • we didn't really have a good understanding of SM300 motors

3. Identify IOCs we can contribute back to EPICS Community

   • We have quite a collection of IOCs now. Some could be of value to others.
   • Identify IOCs that would be useful to others. Work out the best way to package them up for distribution. DO to organise a meeting to discuss further.

4. Deployment

   • We left deployment until the last minute, mainly because we were waiting for GEM & SANDALS tickets to be completed.
   • We should recognise that we don't have to delay the release for late tickets. We can proceed with the release and hot-fix any instruments that require the late tickets.

5. Build Speed

   • Full, clean builds on Windows take rather a long time - up to 8h.
   • We should investigate ways to speed up the build. Consider as a topic for stand-down day.

6. Identify Motion Control Tickets

   • create a motion control label in GitHub to allow the Motion Control group to easily identify tickets of interest to them - DO to action

7. Shadow Instrument Scientists

   • Individual members should shadow instrument scientists for time-to-time - to watch them using IBEX and to learn how it could be improved, to better support the things they need to do.