17
$\begingroup$

Scenario

  1. A post is closed.
  2. A user votes to delete that closed post.
  3. The post is reopened without ever having been deleted.
  4. The post is reclosed.

In this situation, the user who originally voted to delete the question cannot cast a subsequent deletion vote.

Discussion

SE recently changed the way in which delete votes could be cast. Historically, if a user voted to delete a question and that delete vote was "cleared" (either through the post being reopened, or the post being deleted), that user could cast another vote to delete the question (assuming that it was reclosed or undeleted).

After the change, a user can only vote to delete a question once. Moreover, unlike close and reopen votes, delete and undelete votes do not age away, thus a user cannot vote to delete a post for a second time ever.

After some discussion with others, I get the impression that the current behaviour is unexpected (from the point of view of a user). The informal consensus appears to be that if a delete vote doesn't contribute to action being taken, then the user who cast the vote should be permitted to cast a second vote (or, alternatively, that if a question is reopened and then reclosed, the existing delete votes should remain).

The purpose of this question is (a) to get clarity from SE about the intended behaviour is (is this a bug?), and (b) to discuss, as a community, what we believe the correct behaviour should be.

$\endgroup$
12
  • 4
    $\begingroup$ (Half of this (a) would make a better question on meta.SE, since you are seeking clarification from SE. ) But I see no problem with respect to the second (b) aim of yours, for community discussion posted here. $\endgroup$
    – amWhy
    Dec 6 '21 at 19:27
  • 1
    $\begingroup$ I definitely agree with the second to last paragraph, though I was not consulted among those you had discussions with. The SE change was certainly unexpected (from my point of view, though I was not privy to the discussion you held). And given the "bug tag", we might get one SE person commenting, when they have time. So while part a would be best addressed on meta.se, I have no problem with your posting it here. So add my name to the "impression" and what you refer to as an "informal consensus". $\endgroup$
    – amWhy
    Dec 6 '21 at 19:35
  • 4
    $\begingroup$ This is inconsistent with the behavior of close votes that get cleared: those can be recast after a waiting period. $\endgroup$
    – gparyani
    Dec 6 '21 at 21:31
  • 4
    $\begingroup$ @gparyani It appears to be inconsistent, but seems to be consistent in the sense that a user may cast a second close (or reopen) vote after such a vote has "aged away". After a certain (slightly variable) period of time, a close vote ages away, and can be recast. Delete votes never age away, so the existing mechanism does not allow users to cast second delete votes. $\endgroup$
    – Xander Henderson Mod
    Dec 6 '21 at 23:14
  • 3
    $\begingroup$ Strictly speaking, though, if the vote is cleared for any reason and didn't contribute to performing an action, re-voting is allowed (with restrictions) for close votes but not for delete votes. $\endgroup$
    – gparyani
    Dec 6 '21 at 23:17
  • 1
    $\begingroup$ Also worth noting that the original community-imposed rule on Stack Overflow that led to this system rule specifically exempted delete votes cleared as a result of question reopening, so I'm not sure why the team opted not to implement it in the system rule. $\endgroup$
    – gparyani
    Dec 6 '21 at 23:18
  • 4
    $\begingroup$ +1 This need to be corrected since otherwise attempts to workaround it would be extremely cumbersome, which would greatly inhibit site quality control efforts. $\endgroup$ Dec 7 '21 at 1:13
  • $\begingroup$ I am not sure, but I think I was warned that I can't delete my post again once I deleted and undeleted it. This may be because too much post deletion can pose a problem, not sure what that was but I know one can't just post and delete it. $\endgroup$
    – soupless
    Dec 7 '21 at 5:37
  • 8
    $\begingroup$ After some internal discussion, we agree with the points here - the intention of the single vote to delete is to prevent delete/undelete wars, so in cases where the deletion doesn't go through, privileged users should be allowed to recast it, other restrictions allowing. I don't have a timeline for when this can be implemented but it's on our backlog to work on in the future. $\endgroup$
    – Catija StaffMod
    Dec 10 '21 at 22:07
  • $\begingroup$ Thanks so much, @Catija!! $\endgroup$
    – amWhy
    Dec 10 '21 at 22:33
  • 4
    $\begingroup$ @Catija I'm in favor of this. I'd suggest making the change a bit more expansive. I'd say that for all of delete, undelete, close, and reopen, the vote should be permitted to be recast as long as the vote didn't participate in changing the state of the post. This would include things like being able to recast retracted votes. It would be reasonable to enforce a wait between when when it's retracted and when it can be recast. How long that delay should be is something to consider, and may be desired to be variable based on how long it's been since the vote was initially cast. $\endgroup$
    – Makyen
    Dec 10 '21 at 23:13
  • 1
    $\begingroup$ For example, if the user initially voted a couple/few minutes ago, then it would be reasonable to allow them to recast their vote after a few seconds. If it's been hours or days, then substantially longer delays could be appropriate. Perhaps with close votes, if the close-vote has been active for 6 hours, then they have to wait the 14 days which they would have had to wait had the close-vote aged away at that time. $\endgroup$
    – Makyen
    Dec 10 '21 at 23:13

You must log in to answer this question.

Browse other questions tagged .