The issues being:
- mUzima Software artefacts (like android application APK, OMODs, etc) are being generated manually .
- Artefacts are not easily discoverable. Currently they are uploaded on dropbox and shared either via email or jira comments. Due to several APKs being shared, it is hard to know which APK to use for testing.
- Existing automated tests are not being run. This could be for various reasons like they are not stable, take long time to run, difficult to write new tests, difficult to maintain existing tests, do not provide enough value, etc.
- Most of the testing is manual. There are a few unit tests but they are not being maintained, some of them are failing.