Functional testing, why just black box?

Question

I was reading on Wiki and few other places about functional testing and I do not get one thing - why is it considered as purely black box testing? When I test functionality, I often rely on knowledge of the implementation or at least things like boundary values (or data types in DB etc.). Or am I missing something here?

Answer 1

This is just one aspect of testing, while there are also others relying on having knowledge about internal structure of your program.

Advantage of this approach is ensuring objectivity during testing. Knowing internal structure of program, tester often (un)intentionally avoids corner cases, or over-concentrates on testing cases where he thinks they might be.

This is why functional testing is often conducted by third party not involved in development process. Of course, you are free to let that person know about boundary values as it does not really reveal internal structure.

The main idea here is to concentrate on creating thorough test cases and comparing acquired and expected output, while not bothering thinking how program internally works it out.