Should We Just... Delete It? - João Proença

Description:

Have you ever looked at a failing automated test and asked yourself... should I just delete it? I've asked myself this question numerous times, I've trained myself to do so, because I understand its importance. However, I'm often left baffled by colleagues and other testers who reject the idea of deleting a test. Why do they find it such a scary concept?

So the result is that we continuously look and review the same tests, not knowing if they’re mitigating any risks, or worse, not knowing what they are testing anymore. But if you're like me, you've probably looked at some failing tests before that left you thinking "why does this even exist?!". That trigger is one you shouldn't ignore.

In this talk, I'm going to share my experiences of listening to this trigger, but more importantly, I'll explain the actions I take. You'll learn how to analyze the full lifecycle of an automated test to truly understand its value. We'll talk about the total cost of ownership of a test, and how it is a key analysis factor when attempting to reduce feedback loops. I’ll bring some stories from the company I work for to support this. After all, we’ve gone from automated regression environments running thousands of tests overnight to a CI/CD reality. 

I love my delete key, I hope to share the love!

Takeaways

• Understand all the costs of a test, from its development to when it’s run and maintained.
• Learn key criteria to apply when deciding to delete a test or refactor it.
• Acknowledge that as the number of automated tests grows for your system, the more you need to prioritize and understand which are the most valuable ones.