13

I came across this bug report that was closed after it has been tagged . It looks like a perfectly fine bug report to me, but somehow it was closed as off-topic.

Take a look at the close reason:

The problem described here can no longer be reproduced. Changes to the system or to the circumstances affecting the asker have rendered it obsolete. If you encounter a similar problem, please post a new question.

Great. Now let's go ahead and close all these 4206 questions on meta tagged with and .

Well...

Back in the days when "too localized" was still around, people questioned whether we should close all these completed bug reports with that reason. Shog9 explained that these questions are on-topic and should not be closed.

So what kinds of bug reports are accurately described by that close reason? Well, bug reports that involve a situation where the OP manages to fix themselves by clearing the cache or disabling some faulty extensions, or old bugs on some of the site features that are never fixed but are completely replaced, like bugs on the old top-bar that can no longer be reproduced because of a change to the system (there's the new top-bar).

Should we stop closing the bug reports that are simply fixed? Or should we close them because we don't want people to answer them?

  • 3
    Agreed, but old reports for functionality that no longer exists (replaced by different functionality, etc.) are fair game as far as I am concerned. – Martijn Pieters Feb 1 '14 at 14:52
  • 1
    @MartijnPieters Yeah, I have mentioned that in my post already. – Antony Feb 1 '14 at 14:55
  • I don't get it. You want users to continue posting to irrelevant bug reports? – Uphill Luge Feb 1 '14 at 15:12
  • That's just wrong, I feel. It's either status-norepro or status-completed, but not the latter along with that close reason. – Arjan Feb 1 '14 at 16:56
9

I generally reserve that close reason for questions about features which have since been completely removed or replaced, and the original report is now obsolete. As in, the feature no longer exists, so the bug report isn't exactly useful to anyone. Whether or not it was marked as completed or declined or whatever is pretty much irrelevant.

For example, take flag weight. The feature has since been removed. All the support questions and bug reports talking about flag weight are now pretty much useless to anyone here, and should be closed as such. The only thing that might be considered useful in the future would be a feature request to remove it, because that would explain to people who were aware of the feature why it was removed.

I don't see any reason whatsoever to go around closing all questions which are tagged as status-completed, so long as the feature they relate to is still a part of the system.

  • I thought flag weight was still in there for mods or the Team to see, and used in the backend calculations for the quantity of flags a user got. – Lance Roberts Feb 2 '14 at 3:57
  • 1
    @Lance Nope, it has now been completely dropped from the database and literally does not exist anymore. No more confusion. :) – animuson Feb 2 '14 at 3:57
  • Thanks, good to know. – Lance Roberts Feb 2 '14 at 3:58
8

Meta is a bug report/ticketing system shoe-horned into a Q&A site. We know it's not perfect. I think the only thing objectionable about these closures is that they're under the heading "off-topic."

This description seems perfectly reasonable to me:

The problem described here can no longer be reproduced. Changes to the system or to the circumstances affecting the asker have rendered it obsolete. If you encounter a similar problem, please post a new question.

That is accurate, and when taken together with the and tags, I think it's pretty clear what the status of the post is. It was a bug, but now it's fixed.

I think the purpose of closing these questions is that we don't want new answers on them. A new post without the tag will be less confusing (and it will get more attention) if a bug manifests again, and the new post can always be linked to the old one if they really are the same.

  • 1
    Closing are for questions that are not a good fit on the site, and so we want to prevent people from answering it because we don't want to encourage irrelevant questions. Experienced users know very well that they should not answer these completed bug reports under most circumstances. If you are afraid that new users would answer them, protect the questions. Each feature has a meaning of its own. Maybe you would like to read the answers of this question as well. – Antony Feb 1 '14 at 15:22
  • @Antony There's a canned close reason specifically for this, so no, closing is not only for questions that are not a good fit on the site. As I said, this is Meta. All the features of a Q&A site have been repurposed. – Bill the Lizard Feb 1 '14 at 15:26
  • 3
    @Antony ...and yes, those questions should absolutely stick around, but I don't see any reason that they need to stay open. – Bill the Lizard Feb 1 '14 at 15:29
  • So should we actively close these 4206 questions? Having tagged appropriately and closed at the same time is a bit redundant. And if some of them are closed while others are not, it starts to confuse people. – Antony Feb 1 '14 at 15:43
  • @Antony I'll close individual posts when I see activity on them that should have been a new question, but I don't see any reason to go close 4k questions that no one is really paying attention to. I don't think people are confused. – Bill the Lizard Feb 1 '14 at 15:48
  • 4
    I disagree. I'd use status-completed when something has been changed, and status-norepro if it cannot be reproduced. Also, I feel none should be closed as that prevents people from posting detailed additional information when the bug shows again. (cc @Antony.) – Arjan Feb 1 '14 at 16:53
4
+50

(TL;DR; at the bottom).

Technically speaking, if a bug has been fixed then "the issue can no longer be reproduced", however it's a bit silly to mark it "no-repro" - of course it cannot be repro'ed anymore, we've fixed it, "removed the bug"!

By that logic, whenever a bug is fixed we should also remove the tag because there is no longer a bug, it has been fixed.

A question being "completed" is denoted in a few possible ways:
By it receiving an adequate answer, tag, tag, or other "final outcome" tags, or of course it being on-hold/closed.
However, once a question is "completed" and has received an adequate final outcome, we shouldn't change it to another "completed" status based on the previous completed reason.
We'll just be chasing our tails.

That is, "status-completed" is job done, finished, nothing to see here, leave it now.

To take a bug question which has been completed and tagged "status-completed", and go on and close as "no-repro" because the bug no longer exists is just daft, and a waste of time.
It was already marked completed by the big red tag.

I'm not being pedantic...

...but where do we draw the line?

In a question with both and , should we remove the "feature-request" tag because there is no feature request as it's been implemented?
The feature now exists as it was completed/implemented, and a request for a feature which already exists is not welcomed.

Is there a close reason for that? As that's exactly what we're doing by closing "status-completed" bug reports.

I'm dipping my toe in the pedantic pond, but it's a valid point.
We shouldn't just keep on re-labeling things based on a previous label and changing the question state when the previous state was acceptably a "completed" state.

If a bug report is tagged , then it is not a bug, it's how it was designed. Should we remove tag, because it's not a bug?
And/or close as "no-repro" as there was no bug in the first place?

Or, is the big red an "adequate" completed and final status on the question?

Same logic, on topic

And this dodgy logic is the same with questions with both and tags.
"" means bug fixed, this is done, finished with.

However by closing as "no-repro" as well as "status-completed", we are stating two different things on the same question:

  1. This bug reported was valid, repro'ed, and now fixed (status-completed)
  2. This bug report is invalid, the bug cannot be reproduced, off-topic/on-hold as no-repro

Which one is it? Completed, or not a bug (norepro)?
It surely cannot be both as we're mixing messages here.

The two are different question states

For a bug to have been fixed, it was at some point:

For a bug report to be marked as "can no longer be reproduced", it should never have been any of the above, because it must have been found and reproduced in order to be fixed.
And while it can't be reproduced now, that is because we fixed it, so leave it at "bug found, fixed/status-completed".

"No longer be reproduced" should be:

  • Bug reported
  • Bug not identified
  • Bug not reproduced
  • Question marked as "can no longer be reproduced"

Clearly the logic is there to close a question as "no longer reproduced" when a bug has been fixed, as technically it cannot be reproduced any more.

But, there was a bug, it was identified and reproduced, and fixed.
So it's just not sane or sensible to now label it with:

Put on hold as off-topic

This question appears to be off-topic. The users who voted to close gave this specific reason:

The problem described here can no longer be reproduced. Changes to the system or to the circumstances affecting the asker have rendered it obsolete.

Yes, changes to the system - it was fixed!


EDIT

@Roombatron5000 made a good point which I agree with and add as an additional note.

If advancing the completed status of "status-completed" serves a logical purposes then I agree with it.

For example this question is "status-completed", but is about which is no longer an active feature.
As such, while was a solid enough final to the question, now the Stack Egg feature is gone it's logical to advance the question's "final status" to "closed as no longer reproduced".

TL;DR;

If a "final outcome" to a question is accurate and substantial enough, such as , or has an answer, then there is no need to advance to another "final outcome" such as close as no-repro.

However, if there is a good reason to advance, for example a question about , then advancing from the final outcome of "status-completed" to "norepro" is logical when Stack Egg is no longer active.

However, advancing a bug report to "no repro" is not a valid reason, because the "status-completed" is the valid and final outcome.

3

The problem isn't so much that they get closed, but if people start deleting them. With the format we have, once they're fixed they should probably be closed, but they shouldn't be deleted. We've found plenty of fixed bugs that users are finding aren't completely fixed. They should also be left as a reference so when someone is researching a new bug, they may find a clue as to what is going wrong in an old one.

You must log in to answer this question.

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