11

The story starts on a (this) Saturday afternoon, finding myself trying to print something. Out of the printer comes a pretty blue piece of paper. I am pretty sure the blue is not coming from one of my creative burbs, so, I am looking up on AU what could be the cause.

And: yes! it is an existing problem apparently, caused by a... bug, which is, I understand, fixable for as long as the "real" fix is not available. The question is however closed, and the only answer does not work for me.

Of course bug reports need to be done, but numerous people would be happy if they could fix now, and printing is a pretty essential thing.

Most likely, I will find my way around in the next 15 minutes or so, but could we, for the sake of being a useful site, not close the question if an answer can, and most likely will, help the undoubtedly many people in the same situation?

Related, maybe even a dupe, but just to refresh the thought: Please reconsider closing bug-related questions if a workaround can be produced

  • 3
    I agree, workaround answers would be very useful, in addition to having a properly filed bug report to eventually get a proper fix from the developers. However, as a close voter, it is hard to decide if such a workaround might exist, and there is general consensus that we should not just collect bug reports here as they don't get seen by the relevant people. Therefore we close them usually, as we don't want to be a "bug reporting service" either that takes your question and pastes it on Launchpad. So, how would you recommend we should decide when we see a bug question in the closing queue? – Byte Commander Feb 23 at 16:59
  • 1
    @ByteCommander yeah, that's a good one. In general, if a bug on really essential functionality occurs, like this, someone finds a way to at least temporarily revert updates and hold packages, or finds a way to work around the issue. This is such a bug, and as fresh as it gets. My suggestion then is not to close immediately, simply just because it's a bug. – Jacob Vlijm Feb 23 at 17:05
  • @JacobVlijm Now, what's the original question to which you refer here ? – Sergiy Kolodyazhnyy Mar 1 at 20:37
  • @SergiyKolodyazhnyy it's the linked one . – Jacob Vlijm Mar 1 at 22:06
12

Problem is that close-as-bug option is misunderstood a lot and we need to pay attention to what the question asks. And I completely agree - we should not close questions that actually ask for help or at least reopen question when a fix/workaround is available.

When a question asks, "Hello, Ubuntu guys, here's an issue. Please fix it" or "I installed Ubuntu and this thing is broken. Devs need to fix this", that's a bug report. Ask Ubuntu is a community site, and despite the fact that some participants are developers or Canonical employees, most users are neither of those two, and Ask Ubuntu has not a lot of word in Ubuntu's development. And that's exactly what the close message means:

"Bug reports and problems specific to development version of Ubuntu should be reported on Launchpad so that developers can see, track and fix these issues."

Don't tell us you found a bug, report it to the developers.


When a question asks, "Hello, my thing is not working properly. How can I deal with it ?" - that's a valid question, which occasionally might turn out to be a well known bug, potentially with a workaround or a fix, or it may even be a duplicate. There's plenty of examples across the site, some of which include:

As for the blue color in the printer, that was me - I hacked your printer to print my favorite colour. In case readers are confused by this statement, please visit this

  • 1
    Haha, I knew it :) – Jacob Vlijm Mar 1 at 22:05
  • 1
    My kind of humor :-) – linuxandria Mar 1 at 22:28
  • 1
    While this sound sensible, the thing is that once the user finds the solution to their issue, they do not report the bug. We want all bugs to be reported, and the close reason was there to allow precisely that. Also, once the bug is fixed the question is non-reproducible. – Braiam Mar 6 at 13:43
  • @Braiam Your assumption is "OP has issue. Issue is an unreported bug. OP gets solution, and doesn't report a bug". What I'm talking about here is "OP has issue. Issue is a reported bug and OP isn't aware of that. We troubleshoot, investigate, and find out there is a bug report. We provide solution within our capability". As for non-reproducible part, to be fair half of this site and whole StackExchange network probably should be closed as outdated and not reproducible. – Sergiy Kolodyazhnyy Mar 6 at 19:22
  • "We provide solution within our capability" and that's the problematic part. It's not always within our capabilities to provide a sensible result. Heck, the devs of apt were forced to implement socks protocol on http because "the internet is actually believing Acquire::socks::Proxy would exist". There are answers on the entire site saying to use that configuration, when it doesn't exist. We couldn't correctly read a manual page, less figure out how to solve a bug. – Braiam Mar 7 at 14:08
  • @Braiam Mhahah, welll apt developers probably could have used a bit more discretion there, but that makes for a good laugh, you humored me. Though I'm not sure what does that have to do with anything here even remotely, but good laugh nonetheless – Sergiy Kolodyazhnyy Mar 8 at 0:47
8

I feel that we should continue to close questions which are or should be bug reports, but we should apply our policy less strictly to questions which are not bug reports but are related to bugs, or change our policy to allow such questions in at least some cases.

While reviewing recently, I noticed a number of questions that I felt really ought not to have been closed as bug reports being closed as bug reports. This prompted me to read some of the meta discussions about bug reports being off topic, and I found that those discussions in general, though with some differences of opinion, justified closing questions about bugs as off-topic in most cases. I recently talked about this in chat.

I found the policy embodied on meta (as in what actually happened, for the most part, on the site), somewhat stricter than the policy I would personally like us to have, which is to only close questions as bug reports if they were obviously intended as or should definitely be bug reports or feature requests and could not readily be interpreted any other way.

The reason I would like us to have this policy is that I think the only benefits of closing questions about bugs are that

  • it motivates people to report bugs and make feature requests on Launchpad and prevents Ask Ubuntu being perceived as a bug-reporting service
  • it might result in the deletion of obsolete questions

The former seems to me only to apply to questions which obviously should be bug reports or feature requests. I agree completely that we ought to close the question and direct people to the developers/maintainers in such cases, but when someone asks for a workaround for a known bug, there seems no benefit in redirecting the OP to the developers.

Deleting truly obsolete questions (to prevent visitors wasting their time) is certainly a benefit worth having, I think, but it depends heavily on

  • the question actually being obsolete and useless to everyone
  • the question actually being deleted

This is hard to predict. Many bugs are long-lived. I have been using a workaround for one bug for the last 3.5 years. An answer describing such a workaround would be useful for longer, than, say, a configuration option applying to a particular version of a desktop environment... Thinking about this benefit, we have three options

  1. Assume the above criteria will be met (the question will become useless, and closing it will cause it to be deleted), and close all questions about bugs
  2. Ask reviewers to assess whether the above criteria will be met, and close bug-related questions which are not likely to be useful in future (rather subjective)
  3. Give bug-related non-bug report questions the benefit of the doubt, and assume they will probably be useful, so leave them open (unless, perhaps, they have clearly already become obsolete).

I personally favour an err-on-the-side-of-leaving-open version of option 2., or better, option 3., while our current policy seems to be closest to option 1..

My own practice has been to interpret questions as not being bug reports when another interpretation seems better, and to not close bug-related questions that seem to me to be likely to benefit more from being answered than being closed and deleted. For example I recently advocated answering this question instead of closing it.

I think our policy favours closing this type of question, and I would prefer our policy to change to be more generous to such questions, while still closing questions that really should be bug reports or feature requests.

  • "there seems no benefit in redirecting the OP to the developers" actually there is. Developers could offer potential workarounds themselves, or know what workarounds their users are using to assist them diagnosing the problem. Is not the same "here's an unexpected error" than "here's an unexpected error, and how I prevent it". Also, it allows the developer to debug the user and offer patches before applying a fix for the program. – Braiam Mar 6 at 13:45
  • @Braiam I appreciate these points and I think it's important to keep sending people to developers with bugs, but here I'm talking about long-lived bugs that have already been reported and probably already have reported workarounds. Having the workaround here as well just makes the information more findable for people who don't know about Launchpad. And as Eliah said here sometimes devs ask people to move discussions elsewhere. Basically, I don't want good information to be lost. Kinda surprised by the instant accept of the other answer here though. – Zanna Mar 6 at 14:52
5

The general idea was: if asked before it is a duplicate. If it is a bug, not asked before and there is a fix post the fix as an answer (or the workaround), otherwise close it as the bug with a link to the bug report or comment that it needs to be reported. When there is a fix the original q could be re-opened if need be.

  • Sounds reasonable, which in practice means give the Q a day or two/three to get a working fix, especially on essential stuff like this. I got it working again, helped by the comments that were posted on the original Q, and the answer on a second question, posted on the exact same issue, one day later. – Jacob Vlijm Feb 23 at 18:41

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .