I work as a developer at a large consulting firm. On my small team, we have 4 developers and 3 functional people (essentially 2 testers). Often we, the developers, will get emails from one of the testers saying something along the lines of<p>"I logged a new defect. It's occurring in this environment... but not always. Will you take a look?" These sort of messages drive me crazy.<p>Should the tester be the one responsible for directly pinpointing the error? Or am I just out of line in thinking this? I do sometimes like to live in a dream world.
Ideally the tester should provide exact data to reproduce the problem, but sometimes errors are extremely difficult to reproduce. Timing issues, for example, or problems that for some reason unknown to the tester do not always occur given the same test scenario.<p>If your tester are routinely providing reports that you are confident they should have been able to give more detail on, then explain that to them. Ask them to provide more detail. But if the problems truly are bizarre and hard to reproduce, it might be too much to ask of them. Really depends on your exact working situation.
Sometime QA cannot re-produce some issues. This is very common in almost every company you go to. Our QA has access to log servers, and they try to provide the exception logs if they are available.