Testing Simplified
I have seen testing teams testing over and over again apps and applications for all type regression and integration testing for the smallest feature change until that project or product is delivered. Stakeholders are always concerned that team will keep up to their committed time frame even in scrum without due consideration to rework.
What are the specifics of end to end testing? Very vague. That’s true that shippable software should be tested in complete sense and be robust but all is perception.
Since this is the sequel to my post “Dealing Continuous Delivery less than ideal”, I must say testing is not untouched. It’s good that some of those testing can be automated but not always. What value are you getting by testing the whole system when you only changed a small part of the functionality?
At least for some automated tests we can do following 2 steps:
- For each user story completed we should write automated tests for it. When those pass, story is done.
- For all subsequent stories we do above and run all the tests to see if anything broke that need any fixes.
The above will reduce the load of manual testing but frequency need manual testing need to be determined.