The point is that when someone asks for code and output, the user with the anomaly either has to show some, or people give the question up as unresolvable. We need the data.
People don't just say 'well multiple teams have written code and gotten output that agrees with what I'm saying.' We need the actual information, not vague reports that a friend of a friend thinks there were test cases.
Oh yeah no totally. I mean, it is indeed difficult in those A/B problem situations where you don't even know enough to ask the proper question. Perhaps SO needs an army of question vetters, code experimenters, and educators?
People don't just say 'well multiple teams have written code and gotten output that agrees with what I'm saying.' We need the actual information, not vague reports that a friend of a friend thinks there were test cases.