A typical workflow of a problem is the one who works on the error, solves it, and the one who discovered the error, the one who decides whether the resolution is acceptable. If so, they close it. If not, they again open the error for further discussion / work / wrangling.
The exception is that the error is a duplicate, often the person who is working on the error understands that it is a duplicate, and can then close the error as a duplicate. Or maybe they resolve it as a duplicate, the novice agrees and closes it.
IIRC, JIRA has a fairly flexible (if complex) workflow, so you can set up any process that you think is appropriate for your team and groups that will send problems.
Edit: I understand that I actually did not pay attention to closing closed questions. In my experience, this often does not happen because people are looking for a system of problems for existing errors that exhibit the same behavior as what they see. And if you're lucky, quite often errors open without any investigation of existing problems.
Having said that, the QA or field man will go βI remember this mistake. Damn, they said it was fixedβ some time after its owner closed it. At this point, they can reopen the old error or create a new one, as well as a link to the original. My preference is that there will be a new error and link, rather than reopening. The reason is that the βnewβ problem may exhibit the same behavior, but it may have a completely different reason. This often happens when messages with a generic error message fail.
Glenn McAllister Jan 17 '11 at 10:13
source share