Before we upgrade the live server we will make a fresh copy to a development instance and then upgrade that. We apply a few core-code tweaks to fit with institutional needs (nothing major) and we have several add-ons which need updating and installing too. The upgrade process is neither taxing or easy - it sits somewhere gloriously in the middle.
What we do spend a lot of time on is testing. We have a rather ghastly excel spreadsheet which has some unit tests in (but they're far from perfect) and we run through the test, adding if it passed (or not) and comments, an example is below:
As you can see from the tabs it covers a lot of Moodle but using it is not only tediously dull but the interface Excel offers is TERRIBLE
My question is, as institutions develop large Moodle installations, each customising and tweaking it to their needs how is testing performed?
My vision is QA being performed in a much easier, more straightforward manor with better tests and using a better tool. All I can see so far is the idea of using JIRA - which may lend itself well to this application and perhaps even Unit Tests which can be shared for institutions to adopt, modify and use for their QA. Perhaps even building into a shared QA testing area, open to more people and allowing institutions to work together on testing (although this has complications due to local customisations)
Any advice or comment would be most welcome.
