Suppose one votes to close a question in "main", and the poster asks in "meta" why it was closed. Apparently the existence of that discussion in meta is not included among the "notifications" received by those who closed the question. Since it would be impolite for the latter not to respond if they knew the discussion was happening, this amounts to a bug ('bug' in the informal sense of bad design, not the technical sense of the 'bug' tag which refers to unexpected behavior).

  • 7
    $\begingroup$ I disagree with the down votes. Although Manishearth is right that there is an easy work around, it would be better design to automate this if possible. As such, it is on the borderline between a bug and a feature-request; I have added the second tag. $\endgroup$ – David E Speyer Jun 26 '13 at 19:56
  • 2
    $\begingroup$ @DavidSpeyer I suspect people vote on meta questions here like they do on other SE sites. Namely, a downvote displayes disagreement with the proposal (or opinion or whatever). $\endgroup$ – Tobias Kildetoft Jun 26 '13 at 20:13
  • 1
    $\begingroup$ In the specific case you are talking about, I didn't get any notification when you wrote "@David Roberts", because I hadn't previously commented on the question. Not sure what to do other than hope that people check meta regularly, if they are in the practice of voting to close. $\endgroup$ – David Roberts Jun 26 '13 at 21:59
  • $\begingroup$ @DavidRoberts : Should I take that as expressing agreement with the feature request? $\endgroup$ – Michael Hardy Jun 26 '13 at 22:08
  • $\begingroup$ @DavidRoberts mathoverflow.net/help/whats-meta, the voting section $\endgroup$ – Manishearth Jun 26 '13 at 22:17
  • $\begingroup$ I'm just pointing out what happened in this case. $\endgroup$ – David Roberts Jun 26 '13 at 22:23
  • $\begingroup$ @DavidRoberts sure :) I was just letting you know about the voting differences if you didn't already $\endgroup$ – Manishearth Jun 26 '13 at 22:25
  • 5
    $\begingroup$ This is not really a bug, the system is designed to work this way, so the bug tag seems incorrect for this question. $\endgroup$ – Kaveh Jun 28 '13 at 0:40
  • 1
    $\begingroup$ Speaking of tags, notifications and closed-questions seem reasonable tags for this question. $\endgroup$ – Martin Sleziak Aug 4 '13 at 17:41
  • 2
    $\begingroup$ @Kaveh : Sometimes reasonable people could think it was a mistake to intentionally design the system to work in a certain way, and they could adduce specific reasons for so thinking. Might they not reasonably call something a "bug" in that case even if it was intended? $\endgroup$ – Michael Hardy Aug 4 '13 at 19:42
  • 2
    $\begingroup$ The description of the tag bug states that it is for reporting unexpected behavior: "Indicates you've found an erroneous or unexpected behaviour in the system that needs to be fixed." If something is designed intentionally to work in some way it is not an unexpected behavior and if you report it as bug it will closed and tagged with status-bydesign. This tag has particular use and what some think should be called a bug is irrelevant. When you are requesting a change in the system you should use feature-request. $\endgroup$ – Kaveh Aug 4 '13 at 19:59

In such a case it is customary to leave a comment on the post being discussed linking to the related meta question. This should usually work; and it directs further comments and discussion to meta.

There's little point to notifying a VTCer that the post is being discussed on meta -- s/he can't revoke the vote. As long a you get a good portion of the community discussing it at meta, it doesn't matter who is there.

  • 1
    $\begingroup$ How can that work? If you vote to close a question, do you subsequently get automatically notified of new comments under it? Is so, you're actually saying so would make your meaning clear. If not, then it won't work. $\endgroup$ – Michael Hardy Jun 26 '13 at 22:07
  • 2
    $\begingroup$ @MichaelHardy VTCers can't undo a vote anyway. This informs people who see the post after the meta post is opened up. Which means that they will go there first, instead of continuing a discussion there/voting to close or reopen $\endgroup$ – Manishearth Jun 26 '13 at 22:14
  • 3
    $\begingroup$ Regardless of whether someone who voted to close cam revoke their vote, we have long encouraged them to be willing to explain their reasons to close, and to participate in a discussion if appropriate. Developing and maintaining a good culture of closing questions is important, and hard! Hence making it easy for people voting to close to be aware if what's going on later seems a very worthy goal. $\endgroup$ – Scott Morrison Jun 27 '13 at 5:19
  • 1
    $\begingroup$ @ScottMorrison sure, though as long as you have a strong meta culture, who participates in the discussion actually doesn't matter much. $\endgroup$ – Manishearth Jun 27 '13 at 5:22
  • $\begingroup$ But I do see what you mean; an explanation from the original closer can clear things up. $\endgroup$ – Manishearth Jun 27 '13 at 5:24
  • 1
    $\begingroup$ I disagree with the spirit of this answer. I would certainly want to know if any question I had voted to close on was taken up at meta, so I can reevaluate my vote, and vote to open if necessary. Of course this is not necessary since others can vote to reopen anyway, but in my opinion people should be encouraged to take responsibility for actions such as voting to close. $\endgroup$ – Lucia Oct 17 '15 at 17:34

Clarified suggestion in view of comments (for original see below):

As it is possible to notify various users that interacted with a post in a substantive way (commenters, editors, moderators that closed it) it could make sense to also be able to notify users that closed or reopened a question. This feature might be something that one could add.

To sort of answer my own suggestion. A counter-argument, as @all was removed and only one @ per comment is valid, might be that neither does one wish to have a somewhat bulk-ish notifier for all voters (of a given type) nor does one want to encourage five comments being posted to notify each individually.

Perhaps, it is not that feasible idea in the end. But, at least I learned something via my suggestion. Thanks to Manishearth and Kaveh for information.

While the question grew out of a somewhat complicated situation and thus might not be presented in an optimal way, I think it touches a very valid point. I do not fully share the specific suggestion of automatic notification when a meta-thread is created, it feels a bit complicated, might not always be necessary and is also somewhat limiting (see below).

The main issue it raises in my mind is whether who can be notified via an @ is sufficient/optimal? (I do not yet fully oversee this notification-system so I might say something that is not true, I welcome corrections.)

As far as I understand one can notifiy users that commented and OP is automatically notified. But, I think it could be useful to be able to notify everybody that interacted with a post in a substantive way, that is also users that edited, closed (does one now say, had put it on hold?) or reopened it. Or, to put this differently: users that show up in the revision-history of the post.

I can easily imagine situations where it could be useful to be able to ping somebody that made an edit to a post to get some clarification about a choice they made when doing so, for example. Also to be able to draw the attention from closers or reopeners to some debate could be useful, or also just to ask them specifically in cases of doubts what was the moticatiing reason could be useful.

I cannot really follow Manishearth explication that it does not matter so much who debates and also that people will notice the comment on main or the question on meta. By the same token on could question the usefulness of the entire notification system, since also a follow-up comment can just be noticed or answered by somebody else (and we lived quite well without notifications until very recently on MO). Yet, if one thinks that the possibility to notify is convenient (as I do), then I do not see why one should not also be able to notify a user that edited or closed (in addition to thos that commented).

Thus, I would like to present my slight modification of Michael Hardy's feature request:

Add the possibility to notify those users (via @) that show up in the revision history of a post (that is, users that edited or contributed to a completed closing or reopening).

I hope this would allow to address the problem Michael Hardy raised, and could also be useful in other ways, such as being able to ask for clarifications related to an edit.

  • 1
    $\begingroup$ You can also notify editors, and any moderators who voted to close. $\endgroup$ – Manishearth Jun 28 '13 at 0:22
  • $\begingroup$ My point about "doesn't matter" was that as long as you have a healthy meta community, a good consensus can be reached without the need of the original closevoters (and in my experience at least one closevoter turns up on meta to explain). Regardless of the outcome of the meta post, if enough community members have both discussed and voted for it, this carries much more weight than 5 close votes. $\endgroup$ – Manishearth Jun 28 '13 at 0:25
  • $\begingroup$ @Manishearth: thank you for the information, I did not know one can notify editors. However, how does this work? I just tried to notify François G. Dorais who edited meta.mathoverflow.net/questions/238/double-backslashes there but contrary to the usual notifications there appeared no suggestion when I started to type his name. If one can notify moderators that voted to close, this seems additional reason to also allow this for normal users. $\endgroup$ – user9072 Jun 28 '13 at 0:32
  • $\begingroup$ @Manishearth: On the other matter, in my experience it can be helpful to get first hand information on motivation. It could also be useful to have the possibility to ask briefly for a reason without going through the process of creating a meta thread. $\endgroup$ – user9072 Jun 28 '13 at 0:36
  • $\begingroup$ IIRC, there used to be @all but they removed it long time ago because of misuse: it was creating too many notifications. $\endgroup$ – Kaveh Jun 28 '13 at 0:43
  • 1
    $\begingroup$ @Kaveh: perhaps this is unclear in my suggestion, I did not mean everybody at once, but that it should be possible to notify each of them. $\endgroup$ – user9072 Jun 28 '13 at 0:46
  • $\begingroup$ Btw, discussion of a closing on MO1.0 also worked by someone posting a comment under the question linking to the meta discussion, or was there a notification system? Put another way, is there something that prevents the practice of MO1.0 on MO2.0? $\endgroup$ – Kaveh Jun 28 '13 at 0:48
  • $\begingroup$ @Kaveh: no there was no notification system, as I explicitly mention in my answer. $\endgroup$ – user9072 Jun 28 '13 at 0:53
  • 1
    $\begingroup$ @Manishearth: I found on meta.SO the information that it is to be expected that while editors can be notifed there is no name suggestion, so it is clear to me now. I will modifiy my answer. $\endgroup$ – user9072 Jun 28 '13 at 0:54
  • 1
    $\begingroup$ @quid, sorry, I read your answer but didn't notice it. I guess I am too tired right now to be able to read carefully. $\endgroup$ – Kaveh Jun 28 '13 at 0:54

You must log in to answer this question.

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