52

Megan has written a post on the Stack Overflow blog about changes that are planned/upcoming on the Stack Exchange platform.

I've seen a bunch of discussion about it on chat - but it feels like having it purely on the blog is a little clunky.

So specifically - what does the MSE community feel about the proposed post notices for closed questions? What's missing, or what should be added? What are some of the lessons we've learnt that could be helpful here?

comparison of old vs. new post notices

14
  • 7
    Good to bring this to MSE, but I see we are only focussing on the Unclear close reason. What about the others: Duplicate, off-topic, TB, POB?
    – Luuklag
    Commented Aug 21, 2019 at 15:49
  • 2
    @Luuklag I don't think this is about the close reason specifically but about the elements that make up the post notice itself.
    – Catija
    Commented Aug 21, 2019 at 16:08
  • @Catija, most points made here are ofcourse applicable in a broader context. But I believe there has been some discussion about the wording of the dupe notice in the past, that I cant find right now. Something that would be usefull to take into account here as well.
    – Luuklag
    Commented Aug 21, 2019 at 16:26
  • 3
    @Luuklag You say "I see we are only focussing on the Unclear close reason"... but that's all we've shared in the blog post. It's difficult to give feedback on proposed text y'all haven't seen, no?
    – Catija
    Commented Aug 21, 2019 at 16:33
  • @Catija that is true, but we could still point out what we believe the flaws in the current notices are. Which I imagine could be valuable input in writing the new notices.
    – Luuklag
    Commented Aug 21, 2019 at 16:36
  • 1
    @Luuklag Sure... but I think that should probably be a separate discussion (and there's likely some outstanding discussions of how to improve it bouncing around already). :)
    – Catija
    Commented Aug 21, 2019 at 16:37
  • 1
    Thanks for starting this discussion, @Journey!
    – Meg Risdal
    Commented Aug 22, 2019 at 2:21
  • 6
    Since the change in post notice is going to be applied for entire network, a featured tag so that network users will know some change is happening?
    – Nog Shine
    Commented Aug 22, 2019 at 4:01
  • 1
    Perhaps edit blog and this post's headline to reflect that it is only about the hold/close banner? There are other post-notices, but I see nothing about those here. Commented Aug 25, 2019 at 9:16
  • 4
    @MeganRisdal Can you please post text instead of, or in addition to images of text? Posting images of text is not only inconvenient for everyone (can't search, can't copy-paste), it excludes blind or near-blind people. Commented Aug 29, 2019 at 11:08
  • Hypothetically speaking, I wonder how that would work best with wordpress - which is what SO corp uses on their blog. Commented Aug 29, 2019 at 12:03
  • 2
    I had no idea this post existed. Why isn't it being featured? You would get more eyeballs and ideas if users from different sites knew about this initiative. So far it's only attracted 590 views. Commented Sep 4, 2019 at 2:13
  • 1
    That's a really good question. Featured posts are official. This is not. This is literally a user choosing to bring up a matter of interest to the community on his own. Commented Sep 4, 2019 at 2:55
  • 1
    Sometimes non-official posts still get featured. Have you asked staff whether they might like to feature this or not? It would be good to get feedback from the wider network community as well as main meta regulars. (I've got 13k rep here and I didn't find out about this post until a week later.) Commented Sep 4, 2019 at 7:13

13 Answers 13

33

Update: We've changed the post notices for question askers so that they do not see the usernames of close voters listed - this includes askers with the close/reopen vote privilege.

We considered making it so that post owners with sufficient rep could still see the list, but decided against it primarily for cost/benefit reasons: the way our code is constructed right now, a lot of other code relies on a person being classified as either a post owner or a privileged user, but not both. This separation makes sense in most places, just not in this instance. And because the likelihood of a higher-rep user having their question closed is lower (and they’re more likely to know how to access the voters if necessary), we felt that this was an appropriate tradeoff.

We discussed changing the post notice language for users who do have permissions but decided to leave it as-is, so that still indicates that both the post owner and those with close/reopen vote privileges can see the names. Because post owners are still able to see who voted to close/reopen their question by going into the post history, it would be inaccurate to say that they do not have access to that information. It's merely harder to reach (and lower rep users are much less likely to find it).

We think that the small change we've made is the best first step in solving this problem. It makes it much more difficult for users who might be angry to lash out at voters, while still being honest with voters that the post owner can see who voted on their post. And we're definitely open to feedback on this approach!

The biggest thing I'd suggest changing is who we show the list of close voters in the notice to.

We're trying to prevent that feeling of being ganged up on, right? And while it's not nearly so prominent, the list of people who closed your question is still right there. I'd also ask that those close voters are only shown to those that have the close vote privilege. That prevents newbies from attacking curators, prevents them from feeling ganged up on (at least by closing), and doesn't hide it from those who know enough about the site to cast a close vote.

15
  • 2
    Can we show it publicly in the revision history? I don't have enough reputation to vote to close questions on SO, but I still indirectly participate in the closing process there and would prefer to be able to see the voters. Commented Aug 21, 2019 at 15:07
  • 2
    @Sonic So far as I know, this change isn't affecting the revision history; it's just a face that changes based on the reader.
    – fbueckert
    Commented Aug 21, 2019 at 15:11
  • 1
    the problem with post history is it is displayed too inconsistently to rely on it for any meaningful purpose. Link to it is shown when there are edits but hidden otherwise, gimme a break. If link would be always visible I would wholeheartedly support removal of voters names from close banners (because everyone could easily find them in revisions history when needed)
    – gnat
    Commented Aug 21, 2019 at 15:33
  • 6
    Now that the post history is viewable on all posts, this is in progress.
    – Catija
    Commented Jul 3, 2020 at 23:25
  • I feel like this kind of counter-productive. Showing the names of close voters, proves to new users they are not being targeted. That in-fact 5 users agreed on the closing. Now without this tag a new user could feel they are being targeted or that the process isn't peer reviewed but possibly a system process or done by "diamond" moderators.
    – Skooba
    Commented Jul 22, 2020 at 19:01
  • @kristinalustig, the "Does this answer your question ..." text signed by one user (rather than the Community user) is handy for that one user to add additional information, but it's no good if the user decides that too frequently they must host long debates with the OP; so they delete the auto-comment, and prevents the other voters from editing the main comment. It's even missing the "From Review" text, which has good and bad points, leading new users to think that one person is targeting them; or that replying to the message (which is a question) serves a purpose. -- The former way was 👏.
    – Rob
    Commented Jul 24, 2020 at 5:34
  • 6
    @Skooba we have clear and scary examples of times when users lash out at specific closers because their question was closed or deleted. Up to and including requiring police intervention or having their employers called because of closing someone's question. This isn't OK. I'd much rather a user be angry at the system for closing their question than at a person. On top of that, I'm not actually sure that it does what you think. There are certainly times when a "system" acting on something is better received than a person, or even a group of people.
    – Catija
    Commented Jul 24, 2020 at 14:12
  • 1
    @Rob This change didn't touch the auto comment at all, so I'm not quite sure what you're asking - could you clarify?
    – Catija
    Commented Jul 24, 2020 at 14:16
  • @Catija - Is this what you are asking about: "... handy for that one user to add additional information, but it's no good if the user decides that too frequently they must host long debates with the OP; so they delete the auto-comment ...", maybe read again or which word is unclear ...
    – Rob
    Commented Jul 24, 2020 at 15:11
  • 2
    @Rob I'm sorry... the whole thing is confusing me. What does an auto comment have to do with the close post notice on a question? Are you saying that the first close voter is removing the auto comment?
    – Catija
    Commented Jul 24, 2020 at 15:14
  • @Catija "Are you saying that the first close voter is removing the auto comment?". Yes, sometimes some people delete the auto-comment signed with their name when it says: "Does this answer your question ..." because it invites some people to debate / protest the close vote at length. --- It is generally agreed (by higher rep users) that doing so is unwanted behavior. IF the auto-comment was signed by the Community user (instead of the first voter/flagger) then some people wouldn't flame the closer and some people wouldn't have the ability to delete the helpful information. An improvement.
    – Rob
    Commented Jul 24, 2020 at 15:30
  • 1
    @Rob OK. This is the first time you've phrased it as an improvement. The other comment sounded like you were saying something was wrong with the implementation we went with. This answer, and, thus, the solution was focused on the post notice. The auto comment is deleted when the question is actually closed, Also, adding "from review" would be incorrect in most cases because those are added as the result of a flag or close vote on the post, not from review at all. Having the comment come from community would also strip the person's ability to edit it.
    – Catija
    Commented Jul 24, 2020 at 15:33
  • Also, a month or two ago, we had a discussion about you locking the comments on a post for a period of time. That resulted in the auto-comment being eaten and not reappearing once the lock was released. You said that you'd put it in the job jar for the Devs. --- If the Community user wrote the comment instead then it could show even when locked and not be edited to add unwanted chatter on a comment locked posts.
    – Rob
    Commented Jul 24, 2020 at 15:36
  • @Catija, "Having the comment come from community would also strip the person's ability to edit it. ". Yes it would sit, as-is, and additional thoughts would go into comments from each reviewer, or they would remain silent; and be on with whatever else they were doing.
    – Rob
    Commented Jul 24, 2020 at 15:39
  • Awesome; so glad to see this finally implemented!
    – TylerH
    Commented Jul 28, 2020 at 16:09
83

I don't think it's a good option to remind users of to the option to 'post a new one' in this banner. My gut feeling says it will only lead to repeated questions, worded slightly different but not enough to make it clear. If authors decide to edit their question without much improvement, then at least the question stays closed and requires no further caretaking (except maybe from the Reopen Votes review queue). If they post a new one, that's more work for the community and the author is more likely to hit a question ban.

2
53

There's a few things that could be tweaked.

  • It should be clearer that editing your question is the preferred alternative. Nothing's going to frustrate a new user than multiple closures and a potential question ban. Something like "You could edit your question. If you have a different problem, you may want to ask a new question instead"

  • "On hold" is meant to be a kinder, gentler alternative to "Closed". It might be worth considering integrating it.

  • Folks react to closure badly. Rather than letting new users see who closed the question, we might actually have that as a N reputation privilege. Users don't need to know who closed something as much as how to get things reopened.

16
  • 17
    "On hold" is not only gentler, but also meant to convey that this is (hopefully) a temporary state that can be remedied (taken "off hold") if the prescribed action is taken. "Closed" had a much more final judgement to it. I hope we can restore that expression of helpfulness and optimism back into the closure process. Commented Aug 21, 2019 at 15:39
  • 1
    I wrote a similar answer/complaint on the blog, it's waiting to be approved.
    – Rob
    Commented Aug 21, 2019 at 15:48
  • 2
    @RobertCartaino It being temporary is great; but we have users on Stack Overflow that will vote to delete an on-hold question immediately. If it's temporary, the question should stick around until it's actually closed. Commented Aug 21, 2019 at 16:09
  • 9
    @GeorgeStocker Overzealous deletion feels outside the scope of this particular discussion. If anything, a less final closure notice would seem to at least guide users away from that. Commented Aug 21, 2019 at 16:26
  • 2
    @RobertCartaino my point is that while we say in 'name' that on-hold is different than closure; in practice it's treated the same by site users. Commented Aug 21, 2019 at 16:28
  • 6
    @GeorgeStocker that seems... like unintended behaviour. And that in some cases we ought to be able to salvage a post might be something worth communicating. Commented Aug 21, 2019 at 16:29
  • 1
    Was the "Closed." in the mock-up intended to signal that "On hold" was going away? Or would the real post say "On Hold." until it went through enough time and it moved to "Closed."?
    – tpg2114
    Commented Aug 21, 2019 at 16:42
  • 9
    @GeorgeStocker you seem to be omitting known system features that are currently used to control impact of immediate delete votes - such as requirements for a solid negative score of the question and no less than 3 votes of 20K users and features allowing askers to easily discover, edit and flag their deleted questions etc. Wonder if you do this intentionally or because you think these measures are unimportant or maybe for some other reason
    – gnat
    Commented Aug 21, 2019 at 18:06
  • 3
    ^^^^ I mention this because I have a bunch of various deleted posts at sites where I am under 10K rep and per my experience a question that has been deleted by 20k users looked like most comfortable for the case if I wanted to edit and undelete it - I didn't notice any obstacles at all (as opposed to posts deleted by myself, moderators and roomba scripts which all had some annoying quirks)
    – gnat
    Commented Aug 21, 2019 at 19:35
  • 1
    @gnat somehow I feel like there’s a big difference between an experienced used navigating the system to find their deleted material and a new user. Commented Aug 21, 2019 at 21:19
  • 1
    Not to mention the experience of having your question put on hold and then deleted within hours of posting it. Commented Aug 21, 2019 at 21:24
  • 3
    well. I am confused. From what you say now it looks like the system UX is in the need to change to better guide inexperienced users about features that are there to help them improve. But instead you seem to be blaming trusted users for casting delete votes, why? @GeorgeStocker
    – gnat
    Commented Aug 21, 2019 at 22:00
  • 1
    @gnat credit or blame where it’s due. UX can always be better but the humans deleting things that shouldn’t be deleted is a problem, yes. Commented Aug 21, 2019 at 22:11
  • 2
    @GeorgeStocker now you seem to be blaming trusted users (many of which probably don't have any deleted posts at all) in not knowing about UX difficulties in using features provided for new users (whose experience is also far in the past for trusted users) - do I understand that correctly? Wonder if you would prefer to block the trusted users from their moderation privileges until the UX is made totally perfect for new users (that is, potentially indefinitely)
    – gnat
    Commented Aug 21, 2019 at 22:21
  • 6
    @GeorgeStocker sorry George, but your concern seems off-topic for this discussion. This is only about the UI/UX of the post banner. While the feedback on here might help improving the UX, this question doesn't touch behavioural change (e.g. voting to delete "on-hold" questions with current restriction). Your concern seems affecting a much bigger changes/policies than this current topic, maybe consider opening a separate post to discuss this, either on MSE or MSO? Commented Aug 22, 2019 at 3:16
47

Make it clear that making an edit will cause the post to be reviewed for reopening

The existing notice already asks users to edit their question. Still, despite this, users are unaware of the fact that an edit will push their question into the Reopen Votes queue, and they believe that it will not accomplish much.

I'd recommend adding an extra sentence to the notice to the effect of:

Editing your question now will place it in a queue where users will consider it for potential reopening.


In addition to that, I'd also go a few steps further and make the notice smarter, e.g.

  • "Your question has been edited; it is being reviewed for reopening."
  • "Your question has been reviewed for reopening; however, at this time, the reviewers assessed that your question should not be reopened. See [here] for further steps you can take."
  • "If you'd like this question to be reopened, please edit your question to address this issue, then follow the steps [here]." (in case the 5-day period has already lapsed)
3
  • 10
    It should probably also mention that only the first edit after it was closed that puts it in the queue, since that's not always clear (and could lead to undesirable behavior like trivial edits to "bump" a post).
    – Em C
    Commented Aug 21, 2019 at 17:00
  • 11
    If we're going to do that, we also need to fix it: right now, the first edit by anybody puts it in the queue. If that first edit is by someone proofreading or tag-sorting the post, the post author has just lost the ability to automatically get their question added to the queue.
    – Mark
    Commented Aug 21, 2019 at 22:14
  • 5
    @Mark Yes, I've been trying to raise that for a while, but it's been fairly controversial. See the discussion I started and a negatively-received feature request that suggests a way of solving that discussion. (Do note that I've shied away from suggesting that only asker edits add questions to the queue, as it used to be that way, but was later changed.) Commented Aug 21, 2019 at 22:17
29

A few things I would suggest.

Emphasize steps for improving the post first

If we're going the route of showing different messages to the author and to the public, we should take the opportunity to put improvement steps first.

Your question has been put on hold because other users feel they do not have enough information to give you a satisfying response. In tandem with comments you've received, try to edit your question to be more specific about the problem you're having.


On Hold. This question has been closed because it needs details or clarity. It is not currently accepting answers.

Minimize/Deemphasize use of the word "Closed"

Idiomatically, the word "Closed" means something different on this site than it means on other sites. A new user who sees their post closed will feel like they've lost the chance to ask the question they want; "On Hold" at least implies that there's still a chance to improve. Yes, the text following the statement should clarify things, that closure is intended to be a temporary state, but we should still lead with a sentiment that more holistically represents how we want the user to feel.

As I showed in my example suggestion, I'm not saying we should completely eliminate use of the word (since the goal is to be honest with the user about what has happened to their post), but it would be good to lead with other words that have different connotations first.

4
  • 1
    The point Xirema makes on "closed" can't be emphasized enough, IMO, in terms of how it is received. The SE cogniscenti know what it means, but new users receive it differently. Commented Aug 21, 2019 at 16:01
  • try to edit your question to be more specific about the problem you're having. What if this gets focused on the expected answers? So to prevent "I already told you everything about my problem" they now have to think about what answer they expect. Something like try to edit your question to be more specific about the type of answers you expect or better variations..
    – rene
    Commented Aug 21, 2019 at 19:09
  • @rene I agree that for some questions that get closed that's a valid response; but I'm skeptical that it's the right kind of "generic" response. Most of the time I see these kinds of closures, it's because the problem itself is still very vague or undefined, not because the OP has inadequately explained what kind of solution they're looking for.
    – Xirema
    Commented Aug 21, 2019 at 19:22
  • Yeah, it is kind of hard to explain in a sentence what Jon Skeet needed a whole blog for ...
    – rene
    Commented Aug 21, 2019 at 19:30
24

Make it clear to outside visitors why new answers can't be added

On a closed post, the only indications as to why answers can't be added are at the top in the title, and in between the question and its answers. There's no indication near where the answer box is supposed to be, which means that on questions with lots of existing answers, it requires quite a bit of scrolling to determine why it can't be answered.

Additionally, new users may not understand the meaning of "closed", "on hold", or "marked as duplicate" and the resulting cause, that new answers can't be added.

According to a famous micro-study of new users:

Why can't I answer?

Okay, my test group weren't meant to be posting answers. Still, they identified this as another weird thing. Having just read that banner, and seeing "anyone can post an answer", the next thing many of them did was click on a question to see if its topic matched their question. When they scrolled down to read the answers, there weren't any, and they discovered that they couldn't post an answer either.

"What? But it said anyone could post an answer - how am I meant to do that?"

Of course, the question they're looking at has been closed. This raises several points:

  • they didn't notice [on hold] in the title (or didn't place any significance on it)
  • they scrolled past the close message
  • there was no way to tell why they couldn't answer

When they did notice the close reason, it was generally fairly good at explaining what had happened, so that much is working. Perhaps we just need to emphasise [sic] the reason, and add a note to the bottom of the page in place of the answer controls explaining why the controls aren't present.

I suggest adding a notice to the bottom of closed questions, to the effect of:

This question has been closed; no new answers are being accepted.

Of course, the micro-study isn't meant to be an exhaustive look at what all new users think; it just raises some salient points as to how they might think, so I quoted it here.

7
  • 1
    That so-called micro-study is full of holes. It is totally unscientific, it lasted exactly one afternoon. We don't know how many people were in the test group. We don't know how many were non-native speakers. We don't know how motivated these participants were (not very, judging by the dismal results). Commented Sep 4, 2019 at 1:58
  • 1
    @Mari-LouA Imagine looking at a really long closed post with lots of answers. Now imagine wanting to add another answer there, only to find that there is no answer box. Where is the indication that I can't add an answer? There literally is none where the answer box is supposed to be. Commented Sep 4, 2019 at 3:04
  • 1
    Then say that without citing a bogus study. Commented Sep 4, 2019 at 3:06
  • 1
    @Mari-LouA I disagree with the assertion that it's bogus. It's a micro-study, not a complete observational study, so one afternoon is OK for it. The number of people in the test group is clearly mentioned. Also, (at the very least for the specific quote I cited here), it doesn't matter if one's English level is high or medium. Finally, it also goes without saying that the study's results aren't to be concluded as fact, but only used to make potential indications. What I'm proposing here is for one of those potential indications to be implemented, to see if it actually makes a difference. Commented Sep 4, 2019 at 3:13
  • 2
    I wonder if referring to the "famous microstudy" as a "straw poll" may lend clarity
    – Aibobot
    Commented Sep 4, 2019 at 3:58
  • 2
    Hi, author of said study here. It was never intended to be a scientific study. It was a bunch of observations borne out of frustration with the system. Make of it what you will, but expecting it to stand up to scientific standards... yeah, that ain't gonna fly.
    – ArtOfCode
    Commented Sep 4, 2019 at 7:45
  • In usability / user experience evaluations, "hallway usability tests" are an established method. You don't need a formal survey with hundreds of interviews or forms, you can eminently expose many basic usability problems with a quick poll among a handful of subjects.
    – tripleee
    Commented Jul 6, 2020 at 5:44
22

Another thing that could be helpful is to provide a status with respect to the reopen process on the close message itself. This would tell the user whether:

  • The question is currently undergoing reopen review.
  • The question underwent reopen review, and the result was Leave Closed (can provide hints to ask in chat and/or on the applicable per-site Meta).
  • The question is not in reopen review, but will enter it if the question is edited.
  • The question is not in reopen review, and will not enter it unless it receives a Vote to Reopen from a 3k+ user.

So, a hypothetical message might look like this:

Closed. This question needs more cowbell. It is not currently accepting answers. Learn more. This question will be placed in reopen review if is edited within the next 6 days and 5 hours.

Private feedback for you: This question is seriously lacking in cowbell. Questions on SNL.SE are expected to have lots of cowbell. Edit the question to add more cowbell. Please see our help information on how to accomplish this.

2
  • 2
    "This question will be placed in reopen review if is edited within the next 6 days and 5 hours." Maybe also add something like "This can only happen once, so make sure that your edit is a substantial improvement and fixes the actual problem with the question". (relevant). Otherwise people might be tempted to just change a word or two, thinking that would get their question reopened. Commented Aug 22, 2019 at 11:36
  • 2
    Some context for us culturally challenged: More Cowbell - SNL Commented Aug 30, 2019 at 0:35
22
+50

Please don't hide detailed information about the reason for closure - from anyone.

For efficient site curation, the list of names is much less important than the reason for closure. On many sites, a question which has been closed may be edited to fit site standards by other users, not necessarily the OP. For this reason, it may be important to have visibility about the reason for closure. (On smaller sites where everyone knows everyone, it might also be useful to be able to ping close-voters, but that's less important.)

With only one example to look at, I can't tell whether "This question needs details or clarity" is code for "closed as unclear what you're asking" or if it's just a generic sentence for a closed question. Please continue to distinguish - including in the publicly visible banner - between different close reasons. It's important to understand exactly why a question has been closed, for anyone who wants to edit to improve it for reopening. In an ideal world, there'd always be a nice helpful comment to explain the exact problem with this specific question - but that's not something we can rely on. Different close reasons are a useful tool to accomplish this objective, and it should still remain abundantly clear which reason was used and what can be done to improve the question.

In other words: WHY are you removing the following useful text from the close banner?

Please clarify your specific problem or add additional details to highlight exactly what you need. As it's currently written, it’s hard to tell exactly what you're asking.

What's replacing it, in the draft image, sounds much more generic and less useful.

21

I'm no UX expert, but I find plain "Learn more" links very meh. The way I see it, hyperlinked text should tell the user what they are going to see if they click it, and especially if they might be a bit put off by the "on hold/closed" banner in the first place. I don't instantly know what "Learn more" links to.

The current banner does that in a simplist, but "at least we try" way: it says you're going to see a help center, some page that will explain you what/how to ask, and something about editing. Great!

My two cents, add some explanation in the new banner. Currently I only see the "Learn more" link - does it go to the help center? The "How to Ask" page? Instead of a dull "Learn more", why not something like...

[...] It is not currently accepting answers. Learn more about the "on hold"1 status and the possible steps to reopen it. (and maybe an additional sentence redirecting to the "How to Ask" page)

That tells you, in a single sentence, that there are "statuses", and that while your question is currently on hold, there's a workflow to reopen it (see also Sonic's answer about the reopening part).


1 Or "closed", if applicable.

13

Two major things, and one remediation to what I feel is a tradeoff we’re going to (be forced to) make.

“Ask another question” is confusing

The presentation needs to significantly de-emphasize “or ask a new [question]”, and:

  • when that link is clicked, have a prominent gate which explicitly instructs the user not to re-ask the same question, and
  • instruct the user that if they want this question answered, they must improve this post.

This is covered in more detail in @Glorfindel’s answer, so I’ll focus this answer on the other item.

“Private feedback for you” should contain more information

Right now, with the exception of the list of close-voters (which I understand is the point of making this public/private separation), all the information about why the user’s question is closed is identical between the public and private view.

The header leads the user to believe he’s getting specific, private feedback about his particular question, but in reality he’s not: he’s getting the same generic rationale as the public does.

We’ve had closures forever and closure banners forever and users still ask “but why was my question closed? The banner says it’s Too Broad but it’s not, it’s quite specific”, etc. The particular suggestions for improvement almost always show up in the comments.

I don’t know what SO has planned for changes to comments, but as I said, there is usually a wealth of specific feedback in the comments which could give more focused guidance to the OP in this “private feedback” channel, and I think it should appear there somehow.

Additionally, if both parties know such feedback is private (gated by reputation threshold or editing badge, etc), then a peaceable dialog is more likely to occur. Maybe we could even obscure the commenter’s usernames until the OP accumulates a “positive question record”.

We might also limit the dialog in some way: the OP and/or each participant is allowed N comments, and after that, no more. Either the question gets improved or it’s swept up in the dustbin of history. No more endless debates about “you need to improve your question in this manner” // “the question is fine! Just stop being a rules lawyer and answer!”.

The user gets his feedback, and he can either improve his question or not, and no one is tied up in endless, useless, enervating debate.

Social glue

If we do this, then I’d say we’d need an automatic link to a chat room under every question, because whatever the talk about this not being a social network, and comments being used only to suggest improvements, that’s absolutely not the only way they are used, and with us being human it never will be.

The comments are where humans come into contact, build relationships, build the community, a sense of a shared mission, that makes this whole thing tick.

We can limit comments proper to feedback on posts, and even limit the number you get per post, but we will still need a way for people to talk, and that invitation needs to happen where people organically meet here even if they didn’t seek out conversation: on Q&A.

4
  • 1
    I agree with you that questions need a chat room, discussion page, or some way other than comments directly under the question for the community to discuss them. I think that is worthy of a feature request independent of this discussion.
    – ColleenV
    Commented Aug 21, 2019 at 18:33
  • De-emphasizing the 'Ask a New Question' and related caveats seem like a really big sticking point to the dialog. Commented Aug 22, 2019 at 9:39
  • 2
    @Pureferret Yes; right now the way it reads is “your question was closed, the easiest way to get answers is ask again”. My concern, probably unfounded but there nevertheless, was this was a deliberate move on SE’s part to “be more welcoming” by rendering closure toothless. They know that downvoting and closures are the main reason people complain about the hostile atmosphere here, and while they can make explicit cases for moderating comments, they know explicitly reducing the content rating and curation system won’t fly, so my worry is they’re looking for subtler ways to achieve that.
    – Dan Bron
    Commented Aug 22, 2019 at 12:40
  • 1
    @Pureferret There’s some tangential evidence for this worry. There was a “project reduplication of deduplication”, which the company allowed the community to imagine was to help automatically (using AI/ML) of likely dupes before they were asked, so we don’t get yet more identical-questions-asked-using-different-words, would would be a boon to askers and a long sought salve for answerers, but it turned out SE commissioned this project to identify “false” duplicates, presumably to reduce closure rate, and increase “welcomingness”.
    – Dan Bron
    Commented Aug 22, 2019 at 12:48
11

We need a clearer path for people to follow post-closure

As someone who closes lots of things, I've noticed that very few people ever try to engage after closure. That's a problem, and it clearly plays into resentment of the community itself. Let's pick on the one major reopenable reason on SO, minreprex minimal reproducible example.

Questions is posted, but doesn't have enough data to answer. Maybe you have comments hinting at what's missing, maybe not. What is a new user supposed to do? We're feeding them the list of people who closed the question but... why? As a long time user, I know I can ping dupehammers and maybe close voters, but only if I type their username out fully. And what does editing do? I want my question reopened, not edited! Nothing to them suggests a path forward. What I'd love to see here is

  1. An opt-out notice for close-voters that would let you know a question you helped to close has been edited and/or sent to the reopen queue.
  2. A single, form-driven, notification for close voters (since we don't allow pings to mere close voters), as filled out by the OP. Fill out a form with comments (don't allow a free-form entry into a comment for this) and the system will notify close voters and post it as a comment from the OP. This is a once-per-close thing, so nobody is can get spammed (can only vote to successfully close a question once, unless you're a mod). Lets people know the OP is at least engaged in the reopen process. Many people will not fill this out, sadly.
  3. A link to the reopen page. In fact, the closed questions page only contains a link to that towards the bottom. Maybe we should merge them, so people understand there's a reopen process.

I think we can do better here. Maybe reopening won't solve all the problems, but it might be a step in the right direction

2
  • For #2, what do you envision being on this form? Sounds kind of like a reverse close-vote dialog but I'm not sure what options would make sense.. or did you mean it'd be a special button somewhere but still free-form text?
    – Em C
    Commented Aug 21, 2019 at 21:52
  • I'm thinking that this could be a comment-like entry in the new closed dialogue. Something to request clarification, possibly after an edit.
    – Machavity
    Commented Aug 21, 2019 at 22:12
10

First, thank you for finally working on improving guidance around closed questions. The site is currently very bad at guiding askers towards turning an unanswerable question into an answerable one.

Distinguishing guidance for close voters and guidance for the asker is absolutely a step in the right direction.

I don't understand why the guidance to the asker has become less informative. It should be the other way round! Start with the current text of the existing close reason. The problem with current close reason text is that it isn't informative enough, not that it's too long.

Don't show who closed the question specifically to the asker. That makes it personal. Show who closed the question in a different box, or only as part of the question history. The point is to emphasize why and de-emphasize who. Do show the list of voters to the community, at least to other people with the close privilege: that's important for accountability.

A very important thing that's missing is that questions can be reopened. The guidance for the asker should make it clear that closed questions can be reopened without having to open another page.

I don't think stating “private feedback for you” is a good idea. I'm not even sure that part needs to be hidden from others, and it should certainly not be stated in this way. “Private feedback for you” says that this was written specifically for this particular user, which is wrong: what follows is generic text. You're building an expectation of finely tailored guidance and then going against that expectation by providing a generic blunt hammer.

Here's a better starting point for the guidance for the asker.

This question was closed because it is unclear what kind of answers you expect, or what part of the problem you are asking about. Please clarify your specific problem or add additional details to highlight exactly what you need.

To clarify your question, edit it. Once it has been edited, the question may be reopened.

(where edit is a link that does the same as the “edit” button)

I see that you're removing the “on hold” terminology. At least I hope you are. It was gratuitously confusing to have two names for the same thing (with an extremely subtle distinction — whether the first edit pushes the question to the reopen queue — which it's pointless to invent a word for), and doubly confusing because “on hold” has a meaning which is completely different from its normal meaning.

In a second step, close voters should be able to provide more specific guidance, typically with a link to a meta thread that discusses a specific type of question or issue. This is currently available in a very limited way through off-topic close reasons, but that's not nearly enough, since it isn't available for anything but “off-topic” and the number of such reasons is extremely limited. Currently, close voters can provide such reasons by posting comments, but it's a hassle to maintain a library of guidance text and meta links: this should be built into the system.

-3

I suggest to not hide the list of close voters on small sites.

On small sites, where users "know" each other, the visible list might encourage users to make reviews carefully - after all the whole community will see their name in the list of close voters.

If the list is no longer shown to the community, what will prevent robo-reviewers? There are no audits on the small sites and unfavourable patterns will no longer be visible.

12
  • 6
    If you read the comments on the blog post, there's actually a question about who can see the list and Meg responded that the list of people will be visible to users with the close vote privilege, too. So I think this is covered. There's a balance to reach here... you're right that we want people to be thoughtful about closing but we also don't want to deter people from close voting because they're worried about being blamed for it. I think hiding the voters from people who can't vote is a reasonable middle point.
    – Catija
    Commented Aug 21, 2019 at 15:34
  • 1
    @Catija I think this is too late. Especially on smaller sites 3000 rep seriously limits the number of people who would be able to notice any problems Commented Aug 21, 2019 at 15:38
  • 5
    Close voting is 500 rep on sites with the beta reputation levels. I understand your concern but I think that it's extremely uncommon. Robo reviewers only do harm if there's five of them, which is a bigger problem. If questions are getting closed improperly, what's important is that the questions are getting reopened and if there's a concern of it happening frequently, a flag should be raised for mods to look into it. It doesn't need to be due to a single user reviewing incorrectly. I'm curious - how often do you find reviewers who need to be reported as abusive?
    – Catija
    Commented Aug 21, 2019 at 15:43
  • 1
    @Catija It's also worth noting that while I don't have enough reputation to cast close votes on Stack Overflow, I do indirectly participate in the closing process there in groups like SOCVR, so removing close voters from public view for me would be something I wouldn't like. I'd make it something like 250 or 1000 (respectively, the requirement to close your own questions or that to see vote counts). Commented Aug 21, 2019 at 16:39
  • 1
    @Catija You say "a flag should be raised for mods to look into it." That's exactly the point. If the names are no longer shown, it will no longer be possible see any patterns that are worth to be flagged. As to how often reporting would be necessary: I left the site I cared about, so I won't report anything, but there are a lot of strange pattern that would be worth reporting. Commented Aug 21, 2019 at 16:40
  • 5
    @user36296 No, you don't need to see patterns of users. You need to see patterns of mis-closed questions. If the questions should not be closed, that's the pattern. If the questions should have been closed, seeing the same usernames over and over is not necessary. You can't flag users, you can only flag posts, so focus on the posts, not the users. If the mods see that there's a group of users at fault, they can act on that.
    – Catija
    Commented Aug 21, 2019 at 16:43
  • @sonic Currently the plan is that users can always see the list on their own posts, so the 250 rep level is unnecessary (ignoring that there seems to be a popular opinion that askers shouldn't see the list). If you want to suggest the view votes privilege instead of the close privilege, feel free to do that as an answer.
    – Catija
    Commented Aug 21, 2019 at 16:46
  • @Catija Will everyone be able to see it in the revision history? If that's so, then that's good enough for me and my suggestion would be moot. Commented Aug 21, 2019 at 16:47
  • @SonictheAnonymousHedgehog As far as I'm aware, we're not touching the revision history page.
    – Catija
    Commented Aug 21, 2019 at 16:48
  • 1
    @Catija I can only speak from my experience, but I've seen several users who lash out at close voters. I find it extremely unlikely for that to happen from other users, at least in the comments to the extent where a mod flag is required to deal with the user. If you're going to hide it from someone, why not hide it from users with under 3000 rep? If none of the close voters commented or edited, there's no point in showing who closed it to the user (they can't reach out anyway - imagine 1 rep users. No comments, no chat, no way to ping... What's the point?). Commented Aug 21, 2019 at 17:42
  • @Olivia I'm confused. The current plan is to hide the info from people who don't have the privilege to vote to close questions. That's what I've been saying.
    – Catija
    Commented Aug 21, 2019 at 17:53
  • 3
    @Catija I appear to have missed the part where this was a work in progress. I couldn't find anything in the post on it, so I went off some wrong assumptions. My first point still stands: you said you are planning to always show who closed it to the user, but that can expose the close voters to targeted abuse. Most of those users, again based off my experience and not concrete numbers, requiring the vote close votes privilege from askers could be a good way to avoid some targeted comments. It also won't benefit the user to see it. Commented Aug 21, 2019 at 18:10

You must log in to answer this question.

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