In the past I have attempted to draw distinction shouty bug reports (the things we closed as "not a real question") and a problem that might be a bug. They're clearly related and often the only distinction will be in their wording.
Before I plod any further, we do have to consider how the Ubuntu project works and how this feeds back into the bug tracking system. As I've said before:
Let's consider the following statements:
- Bugs that aren't reported are unlikely to be fixed
- Developers aren't checking Ask Ubuntu for bug reports
You're talking as if Launchpad and Ask Ubuntu users are these two
completely separate user bases. But with the odd exception, we're
Ubuntu users. The conversation isn't really about what's best for
somebody shouting out their bug report, or what's most like another
Stack Exchange site; it's about what's best for Ubuntu.
So if bugs clearly need to be reported, we need to be providing the user with the knowledge on how to push forward with their issue and submit a good bug report. That is often more than just closing with the generic "go to launchpad" error message. We should point out what package to report it under amongst other things.
But what if it has already been reported to some degree?
Something that is happening increasingly often is that people will suggest existing bug reports. They'll drop a link to Launchpad in the comments and move on. Seemingly before anybody can verify anything, the whole thing gets locked down as "This is that bug, go away and read it".
I got fairly angry when I saw this happening once. The bug report was awful but because it existed the reviewers thought that enough to close the question. They didn't read the question (which asked some fairly sensible things) and they didn't answer it.
Since then I've been fairly adamant that the best thing to do (both for Ubuntu and us) is to answer these questions. I'll use my previous post's example answer again:
This looks a lot like it might be Title of the bug report. It is responsible for this, that and the other and exists because of this explanation.
There currently aren't any was of fixing this but subscribe to the bug (link how) and you'll know as soon as it's been dealt with.
A good answer like this is a million times better than a just closing, or commenting and closing for a few reasons:
- You've answered the post without shutting it down.
- You now have a dupe target for the next thousand people who can't/don't/won't use Google.
- You tell the user what to do when they get there rather than just telling them to go away.
- The answer can be maintained by users rather than building up a stack of obsolete comments.
Even if there isn't a workaround, it's a constructive post. If you can give them a brief explanation on what's going on and explain how they help the report or just stay apprised of the bug status, you have helped them. You will probably help the bug too. And again, dupe target.
Before you start throwing rotten vegetables at me:
We have strong precedence of this approach working.
There was a bug in Apt that meant that the merging of various source lists would occasionally break. In 2011 we got a Q and A that dealt with it (as I've prescribed). In 2013 something happened that made it exhibit the issue a lot more but the same bug that was never properly fixed. Since then:
- We've managed to close over 200 questions against it.
- The combined views of all these duped questions lands at around 250,000.
Besides directly helping (which can't be done in every bug), it has shepherded from Google onto Ask Ubuntu (good for our SE overlords) but also moves people onto the tracker so they can take things further.
This isn't a broken window or a slippery slope and I couldn't give a damn how other sites handle this sort of thing, if done well, this approach really delivers.
