How do you decide if a QA problem is a defect?

Question

I worked in several companies as a developer and recently switched to QA automation in a new company. Each company is different, and I have yet to see a way to handle this, which I really like. Too often, the QA says that something is a problem, and the answer is “good, but it will be too difficult and take too much time”, or “this is not a mistake, this is a feature!”.
Has anyone found a reasonable way to determine if QA says something needs to be fixed or not?

Answer 1

As a developer, I know that you always get an error that forces you to swear (under your breath) in QA - but I don’t think that a fix / no fix solution should be provided to the developer - as shown by the excuses you mentioned !! The most humble of programmers is outraged by errors that appear in his / her code, and therefore it may be tempting to give you a hard time. I think that a little friction between testers and developers is a necessary evil (provided that you buy them beer at the end of the day!). “This is not a mistake, this is a feature” - this is a general remark, but sometimes it is valid, so it is probably important that an important person is connected with someone from the business side (if that makes sense in what you do).

In my experience, it’s worth recording material, even if they cannot be fixed right now - you can always assign a sliding scale of priorities and simply fix it to a certain level. Reviewing bugs regularly with the / dev tester together can help.

Answer 2

The way I did this was that one person (Product Manager) was responsible for prioritizing bugs and new features. The Prime Minister decided whether each element was a mistake or a new function based on the following criteria:

• if the software does something that is clearly obviously wrong (i.e. not something that someone wanted or intended), then this is a mistake.
• If something contradicts the design documentation for the software, and it does not have obvious advantages, this is a mistake.
• if the software does what the client (or someone else) does not want, but the behavior matches the project documents, then this is a function request.

PM will discuss every error or feature request with the engineers, as well as customer representatives, and on this basis (as well as common sense and experience) will assign priority to each element. In addition, the engineer will be asked to indicate the approximate time frame for each element, and the PM will use this to plan the next iteration.

In short, the mistake is that the software does not do what the people who planned it did for it, and the request for the function is when someone wants the software to do what was not planned.

Answer 3

The SCRUM methodology provides an answer to this question. The product owner decides that something is a bug that creates an item in the product backlog. Then the element is assigned to iterate depending on its priority.

Answer 4

Functionality errors and user interface errors are easy to find and less debatable. A design error is one that must go through BA and the development team in order to get an opinion. In addition, environmental issues should be monitored separately and should not be included in the category of errors.

Answer 5

There are several ways. some of them:

• Software requirements should be fully described. If you see that some requirements are not met, this is obviously a mistake.

• You see that this requirement is met, but in some non-obvious form. This is also a mistake. But this is a situation where the developer can say "everything is in order" and try to close the error. You could find your opinion (this defect exists):

  • shows how similar things are implemented in one product.
  • an example of how such things work in similar products (for example, gmail can be used as an example of mail hosting, etc.).
  • Ask sales and customer relationship managers what people expect from this feature, how it should work from an end-user perspective.
  • Use industry best practices.

• You see that something works, but can be improved. This is also a defect :). This is similar to point 2, and all the recommendations presented there are also useful for this case.

PS and discuss, discuss, discuss with people from other departments.