Recently, I just tried to compile all of the test runs before the mobile apps team released a new version (performed regression test).
Based on the last 4 previous releases, there was a moment when I’d skip all the test cases to shorten the time for release. However, it’s actually producing more bugs that swarmed the production after we released the version of v3.65.0. After i did the investigation, 5 out of 9 bugs that i found after release were context-dependent to another feature, albeit I’m not skipping the critical/important test cases.
For me, this is a lesson learned that was never meant to happen and provided me with a valuable experience that all of these things are correlated with the fundamental theory of testing principles: pesticide paradox and defect clustering. So, after this, do I need to rigorously test the feature? honestly, it all depends on the context and situation. Exhaustive testing is impossible to do, and even though you did that, there is still a chance that bugs will appear.
But, one thing for sure is that part of testing itself was meant to show the presence of defects — not their absence