Welcome to Codidact Meta!
Codidact Meta is the meta-discussion site for the Codidact community network and the Codidact software. Whether you have bug reports or feature requests, support questions or rule discussions that touch the whole network – this is the site for you.
Bringing bad posts to the attention of curators, not just moderators - flagging vs closure reasons
As things stand, the options for flagging a question seem rather spartan:
Image of flagging dialog
We have:
- two options for actual abusive use of the site software
- an option for duplicates
- "needs author's attention", which to my understanding isn't actually a flag at all and instead redirects to the comment interface
- "other"
On the other hand, the current closure options (for those who have the privilege to vote to close) apparently look like (credit to Monica for the screenshot, from the Discord):
Image of closure dialog
Aside from "duplicate", there's one more option for linking to another Q&A, and then four things that... all seem to me like major sub-categories of "needs author's attention". (Well, except for "off topic", which is normally not fixable.) I want to emphasize here: posts that "need the author's attention", at least in the main Q&A section of a site, inherently should be closed. This is a claim that the question either can't properly be answered, or is not a suitable part of a reference Q&A, for reasons that only the author can address. Therefore, nobody should be trying to write an answer - but we know from the results in other places that people eventually will try.
It's better if we prepare for this well in advance, because the results of being taken off guard are evidently truly awful (and cannot be fixed with even years of post-mortem discussion).
I have three related purposes in mind with this post.
-
I propose that the flagging options should explicitly incorporate all appropriate closure reasons, not just duplicates. (There's nothing really special about duplicates, anyway. The author can even fix that problem in many cases.) It should also make these actually function as flags, in addition to prompting for optional feedback, because they are logical reasons to close the question.
-
I propose that users with the vote-on-holds and curation abilities - not just moderators - should be notified in some way about flags that relate to closure reasons (including duplicates), and I'm looking for suggestions about how this should work. I don't think (for now!) that any number of flags should automatically close or delete a post; but for people who can cast votes, it's clearly better if the rest of us can help focus their attention on what's most likely to need attention. This also entails that the "Why does this post require moderator attention?" header requires rethinking.
-
I'd like to open the floor to reconsider (vs. what others have tried), categories of good reasons to close questions - things that every community reasonably ought to agree is a proper closure reason in principle, in accordance with the vision of community Q&A that Codidact puts forward - so that community curators can be notified of issues, come to consensus, and relieve pressure on moderators. (When I originally posted this, I was unaware that there was already a list - because I naively assumed that the flagging options would reproduce that list. However, I still see room for improvement.)
1 answer
Before I found out about the explicit closure reasons in the interface, this is the proposed set I was working on:
Duplicate
Description: self-explanatory.
Commentary: here is still an open question about what to do with near-duplicates and "related" questions.
Misplaced or off topic
Description: This question is not within the scope of [category] of [site]. It might be better off in a different category or Codidact site, or it may not have a place on Codidact at all.
Commentary: this would encompass what Stack Overflow calls "Not about ..." (i.e., actually off topic), "Seeking recommendations for ...", and "This question belongs on another site". It directly corresponds to the current "off topic" reason, but I think that needs to be broadened. I would expect that moderators and curators can either move questions that aren't otherwise problematic, or else suggest where the author should go to try again.
No clear issue
Description: This question does not clearly identify and ask about a problem that could be experienced by others. It either doesn't make sense at all, is written in a language not supported by the community, or doesn't appear to seek an answer that can be justified objectively; or else it isn't clear why the problem described is actually a problem or that any answer to it could meaningfully help others.
Commentary: this would encompass what Stack Overflow calls "Opinion-based", "Not written in English" and "Not reproducible or was caused by a typo", along with some questions that "Need details or clarity". It's not really captured by existing close reasons. The current "unclear" reason is described much more like what I call "missing crucial information", and "not constructive" only covers the "not reproducible or was caused by a typo" situation (albeit more broadly). While it sounds like this covers a lot of ground, I don't think it's actually useful to differentiate further, most of the time.
Improperly scoped
Description: This question has either too wide or too narrow scope to make a useful reference. If too narrow, it probably focuses on details that are unlikely to be actually relevant to the best solutions to the problem. If too wide, it might be multiple questions disguised as one - for example, asking about a complex task that should be broken down into logical steps that should be handled separately. Alternately, answering it properly may require excessively long answers.
Commentary: this corresponds to "Needs more focus" on Stack Overflow, or "too generic" in the current close reasons (although "too generic" doesn't make any sense to me as a label for this situation). I've broadened it to indicate questions that are too focused as well, because I know from experience that this can be harmful in the long run. In particular, it leads people to ask questions that should be considered duplicates but end up being treated as separate because the answers are too precisely tailored to one version of the same problem.
Missing crucial information
Description: This question cannot be answered in its current form, because an answer would depend on information that only the author can provide. Depending on that information, the question may still be a duplicate or otherwise unsuitable.
Commentary: this corresponds to the other "Needs details or clarity" questions on Stack Overflow, as well as "Needs debugging details". Per the description, it's almost exactly what we currently call "unclear", but I think that label is wrong (see commentary for "No clear issue").
These are honestly pretty close to what we already have, but I think I've made many useful refinements. More importantly, I have not included anything that overlaps "outdated or superseded" - because I don't think such questions should be closed. It would make more sense to be able to add "Outdated" or "Dangerous" reacts to questions, and edit them to link to questions that are about more up-to-date versions of the problem. This is especially important for technical communities - it's important to be able to maintain legacy systems, and the fact that newer systems exist doesn't prevent people from figuring out better ways to maintain the legacy systems.
It comes across like this close reason might have been intended for closing "established" duplicates after the fact, but I think the normal duplicate closure should be used in those cases (and, ideally, authors of answers get notified, so they can migrate and adapt content).
2 comment threads