This is what I currently came up with.
The good:
- Software gets tested, before launch. In most cases by one or multiple people.
- There are automated test frameworks.
- We have requirement documents/ways to describe on how something has to work.
- We have test reporting tools, that can provide statistics on tests.
Based on these things we can already have a fairly well tested website, although there are still some problems, that arise with this.
The bad:
- Cycles are too slow: When doing SCRUM, some features cannot be fully tested yet. Definition of Done is endangered.
- Stuff is only partially available: the feature has been split up in to smaller things to fit into a sprint.
- Communication with developers can sometimes be annoying, some take it personally and get mad at the QA person.
- Testing a full flow can be annoying, if it involves multiple platforms. Example: A user does a signup, receives a verification code via a text message and proceeds in the flow.
- Automated testing can be a challenge. Lots of testers aren’t programmers.
- Having enough unique test data. Trying to come up with ways to restart an automated test. AKA: How to delete data that’s already in the website if a test fails. Without manual intervention.
The reason why I try to sum this up, is that I've got the feeling that at my current job, QA has lots of problems with these things. I wanted to ask if the people over here had any similar experiences?
