I never ask for code to prove a bug unless I can't replicate it myself. (On the other hand, if I submit a bug, I almost always provide code before being asked.)
In this case, though... It seems only 1 person has experienced this, and didn't care enough to even reply back. Closing it due to inactivity doesn't seem to be wrong, to me.
There was no indication that any info was even sent to the user, for all he knows all was fine (and still is) just unresolved. He did his job and reported the bug in detail, oracle needs to do their job and reach out to him. Can it be reproduced? More info needed? Etc. You don't ask for test cases unless its really problematic or the dev can't figure out whats wrong in your opinion.
Most bug-trackers automatically send notices if the user has put in valid contact info. If they didn't, or opted out of the notices, then they'll see the message when they check it for updates.
If they never check it again, time is definitely better spent somewhere else.
If you have a little project and it's getting one or two bug reports a week, it's fine to say that you'll create your own test cases. It's completely different for a large open-source project, where many of the reports are from users/developers who are confused and looking for technical support. With a dozen reports a day and an average of 10 minutes to create a test case, analyze, and reply, we're looking at two hours of time!
I totally agree about the "1 person" part, if several people report a problem it lends credibility to the problem. It may not even be a code bug, but may simply involve clarifying the documentation.
True, but you'll just end up closing real bugs as "didn't submit a test case" if you follow a procedure like MySql. And you'll think you have fewer bugs than you really do.
It's much better to leave them as "unconfirmed" or similar, so that other people can find them and add their own details or test cases. (Of course, I do assume that a project with a lot of "unconfirmed" bugs has no one even looking at the bug database, so it's a double-edged sword.)
> In this case, though... It seems only 1 person has experienced this, and didn't care enough to even reply back. Closing it due to inactivity doesn't seem to be wrong, to me.
The blog post linked to two other reports of the same issue (all closed or ignored), and the author mentioned that he himself had run into it prior to turning up these abandoned reports on their bug-tracker.
And really, given the egregiousness of the symbol exporting issue, it would be nearly impossible for only one person to have run into it.
In this case, though... It seems only 1 person has experienced this, and didn't care enough to even reply back. Closing it due to inactivity doesn't seem to be wrong, to me.