Bug#203299: acknowledged by developer (Closing)
On Sat, Aug 02, 2003 at 06:51:35AM +0800, Dan Jacobson wrote:
> >>>>> "D" == Debian Bug Tracking System <owner@bugs.debian.org> writes:
>
> D> If you are not willing to test fixes for bugs that the developers are not
> D> able to reproduce, please refrain from filing them in the first place.
>
> I don't have the (mental) facilities to test it. I can hardly compile
> a C program. That's why when I think I see a bug I just alert the pros.
And that's why a large percentage of your bugs are closed immediately
because they are bogus. You should only report a bug if you are reasonably
sure it is a real bug; otherwise you are wasting other people's time and
resources.
Similarly for wishlist bugs, don't file a bug for every feature you could
ever imagine. It just clutters the bug tracking system. Think hard about
whether the feature belongs there, is feasible, and whether it is already
covered by another wishlist bug.
Which brings me to duplicate bug reports. It is ESSENTIAL that you check
whether your problem matches an existing bug report before submitting a new
one. Yes, I know that when the bug list is long, this can be
time-consuming, but do it anyway. If you don't, you only shift that burden
to the maintainer, who then must do the same search.
If you had trouble applying the patch and compiling the package, you could
ask for help from someone else, or (gasp) read one of the dozens of how-to
documents which explain the process step-by-step.
> D> I will assume that this bug is caused by user error or misconfiguration.
> D> Closing.
>
> Odd that you got as far as making a source patch for a bug that you
> couldn't reproduce.
I made that change for other reasons, and was interested to see if it fixed
your problem. I can't test this because I cannot reproduce the problem.
> What if I had simply replied with a polite "Thanks!" to your patch. Then
> you perhaps have trustingly incorporated a patch for something caused by
> user error or misconfiguration.
I would never presume a bug to be fixed with only a "Thanks!". If, on the
other hand, you had lied and said that the bug was fixed without testing it,
I think that would be sufficient reason to ignore all future bug reports
from you.
> D> If you are not willing to test fixes for bugs that the developers are not
> D> able to reproduce, please refrain from filing them in the first
> D> place.
>
> But users should still be able to report bugs even if they can't help
> much further.
Occasionally. But an unreproducible bug combined with an unhelpful
submitter is worse than useless; it's a waste of time.
> And fixing things that you can't reproduce is worrisome guessing.
My point exactly. Which is why I asked YOU to test the fix, because YOU
could reproduce the bug and I could not.
--
- mdz
Reply to: