Recently a thought has popped up in my head. Do we still need to regress? It’s like the most ancient aspect of the development process. Every quality assurance team is using it for years. And it seems like the most boring, troublesome, time and effort consuming thing testers are doing on a daily basis. It is practically unrewarding and still, it exists. With so many changes that occurred all around us regression testing is remaining practically untouched? What is the reason for this phenomena?
People are still doing regression testing because it works. Let’s just think about it. As an example, we can take the following scenario. You are all into quality assurance. Here is your new Live Environment right next to you. And, surely, you are desperate of adding something new, cool and shiny into it (whether those would be new features or bug-fixes or stories). And, of course, you have your new-n-shiny process as your weapon of choice like Scrum or Kanban or whatever it is that you prefer.
So what do you do next?
- The QA team has to agree on the Acceptance criteria’s before the process of code-writing may even begin.
- You are informing the Developers and are explaining exactly what you are going to do.
- Then the team gathers around with some coffee and decides to which extent and one what level should the automation take place.
- You are still hanging out with developers. They are showing you some tests they’ve already written as well as some kitten mems and you are adding new ideas (to tests, not kittens) and suggesting new tests that are needed.
- You and the Dev machine are best friends for now. You are testing a lot on it.
- All the automated tests are being run. And what a surprise, they all pass!
- So now you have that happy smile while deploying to your Environment and checking on the functionality. Whether all is working fine and the way it was meant to be.
- You never trust anybody. And that is the correct life-credo, so now for more tests. An exploratory Testing session on the new functionality to be precise.
- Good people do exist. Your testing was reviewed and now you possess has new suggested characters.
- You are adding more automation tests.
- What? There are still issues found after regression testing? How come?
Most likely it all happened due to the wrong goal you were trying to achieve? Don’t fix issues. Fix the reasons causing them to accrue. Sounds like something wise Kung Fu masters with the long fancy mustaches are usually talking about, right?
A great thing to do after all that fuss is sitting down, clearing your mind and thinking on the possible reasons explaining why these issues never came up earlier. After some time spent on solving these puzzles you will actually realize that the process is going smoother. And now you don’t have to spend as much time on regression as before. And thus you will have the opportunity of paying more attention to the cooler stuff.
And maybe, just maybe there will be a day when the regression will transform into ancient history. But until then – regress as good as you can.