15

When a question is closed, anyone can make an edit and submit the question for reopening:

"This edit resolves the original close reason and the question should be considered for reopening."

This can be done an indefinite amount of times, which can lead to situations where authors are making multiple minor edits and wasting reviewers' time in the Reopen Votes queue. In the old system, this could not happen; there was only one opportunity for an edit to trigger the review. That is not ideal either, but I'm sure that two tries should be sufficient (perhaps within a one week window or so). A workaround is to have a moderator lock the question (which I just did), but I'd rather have the system block such attempts.

14
  • Regarding my tags edit, this request is about limiting original poster in something related to review, it's less about the edit itself. So removed the edits tag, adding review tag instead. Aug 23, 2022 at 6:53
  • 1
    As for the request itself, ideally the system would detect such abuse (e.g. three times in a row OP ticked the checkbox, and yet in all three question remained closed) and like in other review actions, give temporary "ban" i.e. won't let them do it. However, those who edit and it leads to reopening, will be able to keep doing it without limit. Aug 23, 2022 at 6:55
  • 2
    Yes, details are open for debate. I was thinking of a limit per question (just like the old situation), not per user.
    – Glorfindel Mod
    Aug 23, 2022 at 7:00
  • 2
    For context, when this feature was first released, the post said "We’ll be launching this feature without any limits – users will be able to submit multiple edits on individual posts. We’ll take a look at the data in a few weeks to see if any abuse of the feature has occurred and take mitigation steps as needed."
    – 41686d6564
    Aug 23, 2022 at 7:00
  • 1
    @ShadowTheKidWizard I don't think it should be limited to OP. They do have the highest incentive to use the feature, but people could do this to others's questions as well.
    – Luuklag
    Aug 23, 2022 at 8:18
  • @Luuklag well, the request here is specific to OP's only, so I just stayed in its trails. :) Aug 23, 2022 at 8:33
  • 1
    I should note that for authors with 250+ rep who can vote to reopen their own question, if it fails to reopen the post in review, they have to wait for the vote to age away plus an additional 14 days before they can vote again. As repeatedly adding one's own question to the queue by this other method is already restricted, nothing needs to be done about that Aug 23, 2022 at 8:35
  • 1
    Rather than preventing a user from submitting the question for reopen review more than X times, perhaps the limitation should be that they can't re-submit for a reopen review unless they made a "substantial" edit, rather than minor cosmetic edits. It's very rare for a question to be appropriately sent to reopen review additional times when the edits are minor. OTOH, it's reasonable to allow submitting for reopen review if the first edit is minor, as it's possible the question was closed inappropriately/inaccurately, so the first trip to reopen review shouldn't have all that high a bar.
    – Makyen
    Aug 23, 2022 at 17:57
  • @Makyen Question is, how would you determine a substantive edit from one that isn't? Checking for character changes may not be ideal, since in at least one site I participate making a minor edit is enough to make the question reopenable (specifically, changing "what should I do" to "how do I do"). Aug 24, 2022 at 7:59
  • 1
    @SonictheAnonymousHedgehog Yes, determining that the edits are "substantive", or, potentially, even just defining what could be a substantive edit, is a significant concern which SE would need to resolve. One possibility is that they could use a similar check as is used for notification to authors of edits by others. Personally, I don't think that's a great definition, but it's one they already use and have the code for.
    – Makyen
    Aug 24, 2022 at 8:07
  • 1
    @SonictheAnonymousHedgehog If a question is so trivially changed from being off-topic to being on-topic, perhaps users reading the question should have made that edit instead of voting to close, even if such a change somewhat changes the thrust of what the questions is asking.
    – Makyen
    Aug 24, 2022 at 8:10
  • @Makyen That's true, but there's always the possibility that even an edit that doesn't trigger it may materially change the acceptability of the question, such as changing a single "what" to "how". Aug 24, 2022 at 8:10
  • @Makyen Voting to close perceptibly takes less time than going ahead and making the edit, so some people who want to just move on will just do so. Also, there's the issue of the author not wanting to ask, in this example, how to do a specific thing but what specific thing to do, in which case one will be making the change against their wishes, and if they want it that way, it should be closed rather than edited. Aug 24, 2022 at 8:12
  • 1
    @Sonic It feels like you're picking a nit here. Yes, there are issues with defining "substantial edit", but it seems much better to allow users to submit for reopen review when the edit is substantial, rather than not permitting them to submit for reopen review at all, regardless of the edit being substantial or not. Arguing against permitting submitting for reopen review when the edit is substantial, because there will be edits which aren't identified as substantial, is effectively nonsensical when the alternative is never permitting the user to submit for review regardless of the edit.
    – Makyen
    Aug 27, 2022 at 17:34

2 Answers 2

2

I think this is a good idea.

I think it's good that we have multiple opportunities to allow a question to be reviewed for reopening by an edit. Perhaps an edit failed to address a close reason or some other close reason also applies to the question, and the question received a comment to that effect or reviewers reviewed the question with that other reason. In that case, one should be able to make another edit to address that feedback and have it reviewed for reopening again.

However, if the question repeatedly makes it through reopen review without reviewers reopening the question, then either the question is fundamentally unsalvageable, the edits fail to address (possibly another rule about) why it's not a good fit, or it's not as clear to reviewers that the question is indeed within the site's rules as it is to the editor making the edit. In those cases, it's better for them to make a post on the site's per-site meta explaining in detail why the question meets site guidelines rather than leaving it up to reviewers to try and make out why. In the first two cases, they can then get a proper detailed explanation as to why the question is fundamentally not a good fit or what other untouched part of the question needs to be edited, and in the third, once the reasons for reopening are clearer, it greatly increases the chance that it will actually get reopened.

My solution is simple: only allow at most three edits with the review box checked after a closure. This limit should apply in total for edits from all users, because I don't believe anything would change in the cases above even if someone else makes an edit, and this will also guard against sockpuppets. Also, of course, this should be reset every time the question gets reopened and re-closed. As I said, multiple reopen reviews through edits aren't a problem, but once the question has had three shots through the review queue, it's better for someone to raise a meta question instead.

On smaller sites where there isn't much crowd on the site meta and review is the only place where discussion on closing/reopening questions takes place, keep in mind that the checkbox isn't the only way to add questions to review: one can also vote to reopen. If the author edits and has 250+ reputation, or another editor has 3,000+ reputation, they can edit and vote to reopen at the same time. This can't be abused because it's already limited, since if the vote fails to reopen the post one has to wait for the vote to age away plus an additional 14 days before voting again.

7
  • 7
    "This can't be abused because it's already limited" > It can still be abused, I know of one post that repeatedly ended up in the review queue without improvements because of the OP of that post kept casting a reopen vote as soon as it was possible. The abuse there is just much less noticable because of the limit on time instead of amount. If any restrictions are going to be made, make them similar and consistent: either an amount limit of three attempts, or no limit on the amount of attempts but on the time that there should be between them (minimum of 14 days).
    – Tinkeringbell Mod
    Aug 23, 2022 at 9:38
  • 1
    @Tinkeringbell That may be true, but much, much fewer people are going to take the time and patience to abuse the system if they have to wait between 18 and 28 days between successive reopen votes than if they can immediately add it to the queue again. There's also a reputation barrier, meaning it can't be exploited at all by totally or mostly new users. Aug 24, 2022 at 8:05
  • The paragraph you cite also applies only to sites with inactive per-site metas but active review queues (which do exist), and so there's no other way to actually have a post reviewed for reopening - if that's limited in the same manner as my proposal, the user will have exhausted every single available method for appeal regardless of the actual acceptability of the question, so it makes sense to keep it open. On other sites, if one seriously thinks the post should be reopened, it's faster for them to make a meta request than to vote themselves or have someone else vote. Aug 24, 2022 at 8:07
  • The argument about abusing the timed delay and putting a post in the queue over and over on sites with review queue activity but no little/no meta activity makes no sense: On sites that do have meta activity, one also doesn't write a post every month asking for the same post to be reopened if the first post was unsuccesful. If one were to do that, they would be called out on it and probably end up being suspended. So why should putting a post in the queues without edits over and over be more valid? It is still abuse, not a 'workaround' for low meta activity.
    – Tinkeringbell Mod
    Aug 24, 2022 at 8:15
  • 1
    As for exhausting the options for appeal: If someone wants to waste their attempts for appeal on posts they don't make edits on/only make insignificant edits on, let them. Again, this is about authors putting their own questions in the queue over and over. Authors aren't the most objective users when it comes to their own posts. If someone else can still put it in the queue, there's no problem with limiting the amount of times an author can put their own post in the queue through any means, as far as I'm concerned.
    – Tinkeringbell Mod
    Aug 24, 2022 at 8:23
  • 1
    To your first reply: I did not mention without edits: they can edit and vote to reopen at the same time.. To your second reply: the request as written asks for the limit of three to be applied to all users, meaning that if the author adds it to the queue once and two other editors do the same (or another one editor does it twice), it will prevent all users from doing so again. Your first comment on this post states that the same restrictions should be applied for editing and voting to reopen. There may be value in restricting voting to reopen for authors, but there isn't for others. Aug 24, 2022 at 8:30
  • My comment said to make things similar and consistent: Either limit the amount of times a post can enter the queue, or limit the time between the times a post can enter the queue. But I don't mind adding that applying a hard limit to the amount of times an edit can put a post in the queue to edits from all users (and not just the author, as the original request states) would be a bad idea as it is too open for abuse. It would allow a nefarious user to block someone from editing and putting their own post in the reopen queue by making three such meaningless edits in short succession.
    – Tinkeringbell Mod
    Aug 24, 2022 at 8:36
1

I understand the argument that people can easily waste reviewer time by repeatedly submitting their question for review with only minor changes. Your proposal is simply to limit the number of times a post can be submitted for review, and if one fails to reopen their post after a certain number of times, then they are mostly out of luck.

However, I do not really like this permanent state of irredeemability of questions, even if one more edit would have made the question good. Maybe the author could make substantial changes that could make the question on-topic after coming back much later with a fresh mind. Every question that could be salvageable deserves another chance... eventually.

I would like to offer an alternate solution: After a question is left closed in review, the question may not be readded to review via the Add to review checkbox for a certain amount of time. The period would increase for every successive left closed in review event that takes place. This way, the problem of people wasting reviewer time by submitting their question for review repeatedly after minor edits would be mitigated. The authors would be still encouraged to make substantial edits because they would know that if they do not, then getting their question reopened will take all the longer. And the longer cooldowns could give people more time to think about how they can improve their question.

I believe this alternate solution would solve or at least help with the abusing the Add to review button with frivolous edits, but would also not cause any questions to be in a 'closed forever' state without external intervention.

You must log in to answer this question.

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