5 ms·
"A great tester is proud to break things." That might (or might not) be true, but this statement is part of why software sucks. Testing software will certainly
by mikestew 9y ago
"A great tester is proud to break things."
That might (or might not) be true, but this statement is part of why software sucks. Testing software will certainly involve breaking things. But a lot of software testing is confirming that the damned thing actually does what it says on the tin. It reminds me of a long-discarded interview question: test a blender. Candidates would throw rocks in it, flip it on/off a million times, turn it on upside down, etc. One thing that I'd guess 90% missed: throw some fruit in it and confirm that it will make a smoothie. A software tester's primary goal is not to break software, it is to give a report on the current state of the product: what works, what doesn't.
The other reason software sucks is because we call the team "QA", as if to imply that if it makes it past those folks, we're done. If it were truly "quality assurance", then a test team would have leeway to say, "there are no unit tests, we're not accepting this build". A QA team could tell the dev team that they will run static analysis tools on the code before a build is released. Or whatever, you get the idea. But software development as it is currently practiced does "QC", or quality control. The QC team has little bearing on the quality, they're just there to make sure it's not a steaming pile, and to give the rest of the team a direction in which to point that finger. If QA teams were allowed to do actual QA, maybe software would suck a little less.
- slededit 9y agoMaking sure it works is a management problem. It's effectively ensuring people are solving the right problem. Testers can point things out but they won't get far saying the premise is wrong.