Hi there,
I am a software tester with 3 years of experience. In your opinion what should come first in application version testing: test new features or test old feature bugfixes? Or it really depends on descriptions and impacts on system of both?
Thanks
Regards
Very much “it depends”, I’d be looking at relative priorities and impacts of the new features and the bugs before making a decision on what to look at first.
I will say both are equally important. If you absolutely have to make a choice, I will say the new feature testing to be more important than the bug testing. Reason being, if the bug was truly a blocker it would have gone in as a hotfix. It going in a normal release means it was not a blocker. When you release a new feature, it is important you are delivering a quality product and is usable.
I am afraid that is very subjective question. It could depend on project organisation while it consists of you decision from where to start. For example, I have autonomy to prioritize tasks and I rather prefer to start from bug fixes task. However, it could be wrong decision if I don’t conceder that completely new development is going to be released too soon in order to I should be focused on testing it.
It is a balance and as mentioned by Adam, it is the priority and the project direction that would drive this. Ultimately, both have to be tested to ensure you meet the exit criteria that is set for the project.
I think it depends upon the impact of the feature or bugfix. For the bug, I think severity and priority come into play as it could something that only impacts a small amount of users or it’s extremely low risk and only touches some of the code so it could be checked by someone else and pushed out without concern.
For the new feature, what is the new feature? For example, if the new feature will generate large amounts of revenue or is time boxed and needs to go out ASAP, that can change what you address and when.
Another dimension you can use if the already mentioned ones are unavailable are what did the developers work on most recent. Or can you provide instant feedback to the developers so they do not have to context switch. I.e. the priority is unclear or not that critical at this stage and the developer just worked on fixing a nasty bug. Go for the bug because if you find anything there they can jump on it straight away.
James Bach introduced a concept that he called shake’n’bake (no idea why) which basically meant you as a tester go to the developers computer and test their stuff in front of them. Two things happen here. They get to see what they can do to do better testing themselves and they can fix any problems right away. In our case something that takes 2 hours now take 10 minutes which is a huge time saver.
Just to be clear, in most cases you should prioritize your work based on importance before you go for avoiding context switching.