When do we say a problem can't be reproduced?

Question

Today I saw a question, noticed it was a problem that I and other people have been having in Ubuntu for years, and wrote an answer. I also checked on my machine that the answer was correct. I would describe what I did as reproducing the problem. A few minutes after I posted my answer, the question was closed as off-topic. In comments, it was suggested that the reason was that the problem could not be reproduced.

This is the subcategory of off-topic that seems to have been intended, though I've gleaned this from comments and this is not what is actually displayed on the question page:

This describes a problem that can't be reproduced that seemingly went away on its own or was only relevant to a very specific period of time. It's off-topic as it's unlikely to help future readers.

---

The OP wanted to edit /etc/fstab:

I am attempting to automount my backup drive on Ubuntu studio 16.04.1.

I type gksu gedit /etc/fstab in terminal and enter my password, but I cannot see the editor.

Can anyone please help?

The OP tried to run the gedit command with gksu and -- as that comment clarifies -- this failed silently. No Gedit window appeared.

I then noticed that the OP had found that Gedit was actually not installed. Immediately I remembered the times I've tried to run nonexistent commands with gksu and got no error message. So I wrote an answer about how gksu doesn't (or at least doesn't always) show any error in this situation. I did mention that the solution in this case was to install Gedit.

I had considered commenting to suggest the OP write their own answer, but because I'm familiar with this problem -- having produced and seen it multiple times over the years -- I went ahead and answered. I also checked that gksu still behaves this way on my machine.

Another dimension of the question is the issue of how root-owned configuration files are best edited. There were comments posted on the question that in effect gave solutions; something like them could have been posted as an answer, but hadn't been. So I also covered an alternative to gksu in my answer. To me, this question is clearly about how to edit files as root (in general) and about gksu (specifically), and was on both counts objectively answerable.

It seems to me that this question is about a problem that gets reproduced all the time -- someone runs a command with gksu, it fails silently, and they don't know what this means or what to do. I believe this question, and my answer (so long as I have managed to write it clearly!), are useful.

---

Any arguments I can think of for why this question should be closed as off-topic -- including the explanations that I have seen in this upvoted comment and others on the question -- seem to apply just as well to most answered questions on the site about solvable problems, or at least to problems where the OP was able to find a solution. I don't know whether I have failed to understand the reasons people have given for why that question should be closed, or if it is simply that I vehemently disagree.

I also don't know what most people on our site think. I've been largely inactive on the site lately, until pretty recently. So I admit that the existing community consensus on what's on topic may have changed, though I haven't found anything here on meta or in the help to suggest that.

Therefore I am putting this question to all of you. What do we mean when we say a problem cannot be reproduced? Is that why this particular question was closed? Should it be reopened?

Answer 1

I think if you just changed the title of the question to "Why does launching an application via gksu not work?" or similar and then keep the body as an example would make the question immediately more generic and reuseable.

Answer 2

I think it's somewhat subjective... but I would like to share my perspective for discussion, even though I am pretty new to the site.

Along with some other folks, I do some cleanup, meaning I spend some of my time on neglected, unanswered, sometimes old and maybe abandoned questions. Mostly, we find that they have good, or good-enough-to-edit-into-shape, neglected answers and we upvote them. On the best days, we can answer the questions ourselves \o/

And sometimes we vote to close them. Most often if we do close them, we close them as "no repro", because the question has an answer from the OP saying something like

• I rebooted, and the problem was gone
• I did a routine system update && upgrade and the problem was gone
• I reinstalled Ubuntu (or I gave up and installed another distro) and the problem was gone
• I did something, can't remember what, but it fixed the problem
• It was my hardware/my ISP/my friends pranking me
• I'd made a typo in sudo apt install foo or similar

We have a multi-level review system and a chatroom (quiet at the moment because of the phase we're in), and we discuss, often, whether questions should be closed as no repro or not. Even if the question has an answer like one of the above examples, we sometimes upvote the answer instead, because we can see ourselves making that mistake or the hardware issue seems reproducible, and sometimes we want to add an answer, because actually, like you in this case, we (ok usually it's not me but one of the smarter ones) have the expertise between us to figure out what action, mysterious to OP though it be, would effectively solve the problem.

The overarching question we are asking ourselves is "will this help anyone else?" and we conclude that it will not, I guess, if it seems to be

• extremely trivial
• extremely specific and unlikely to reoccur anywhere
• a mysterious problem (perhaps due to unclear underlying issues) that has gone away

I don't think any of these apply to the question you mentioned, and I hope I would not have voted to close it (apparently I didn't, thanks Seth!), but I might have considered it fairly trivial and fairly specific and possibly slightly mysterious if it were not for your excellent explanation here, because well damn, obviously you can't launch software you haven't installed and oh hang on wait Gedit is installed in (vanilla) Ubuntu by default (MATE/vim user here; I don't have Gedit either) so what was going on? So while I totally agree with you that the post should not have been closed, I can understand the thought process of the voters.

As I mentioned earlier when talking about our discussions in the cleanup project, sometimes it's our knowledge and experience than stops us from closing a question, and allows us to answer it instead. Your insight made that post a really useful one, now, but that might not have happened; you might not have been there to salvage it.

To some extent then, I think the decision to vote to close a question depends on our ability to answer it! Or on our ability to see value in possible answers; it depends on knowledge and expertise and on attitude too I think. I change my mind about questions all the time based on the awesome insights and knowledge of amazing folks in the community. I guess we should try to err on the side of looking for answers and valuing answers if we can, since answering questions and not closing them is what actually helps people (dupe targets help a lot of course, but only because they have answers!)