-3
$\begingroup$

I liked the answer to this question: An algebra of "integrals" by @anixx

Which had a very detailed table attached. My understanding is the answerer had updated the question some 35 times to generate this table/correct bugs

Some commenters didn't appreciate this since they felt it was polluting the front page but as a person looking for tables like this i am VERY happy to have found someone willing to update that table again and again and again.

So what could we do so the commenters on the answer are happy and I am happy at the same time. It seems like if a question gets edited beyond some threshold Q within a target time period T we should mute notifications/ability to pop up to the top of the front page for edits to THAT particular answer. I think that is the best solution here.

Asking people NOT to continuously update answers/edit an answer for better accuracy/comprehensiveness feels antithetical to the spirit of this site but I do empathize with my fellow users that noise is also not good.

Modified Proposal:

So we select some time period T and threshold Q such that if a particular ANSWER/component of a page gets > Q edits in a time interval T then further edits will not bump the post until the threshold time T has passed. So the goal of this is any interaction with a page after a "while" which is what T indicates, will bump a post but a continuous flurry of edits/conversation should NOT cause a question to stay pinned at the top of everyone's attention.

Concerns that get slightly fixed:

Some other users have asked before about editing -> top of question list creating perverse incentives: How to discourage excessive self-edits?

This doesn't fix the problem if the edits are very spaced out BUT at least we are getting some progress in that direction.

Maybe even just giving the users an option to say "hey you made this edit after this question did NOT have activity in a while, are you interested in bumping this question?" Giving me the option to say "no" would be nice.

$\endgroup$
14
  • 3
    $\begingroup$ If an answer needs editing 35 times, one wonders how to know that last edit really was the last one. Ideally the table is hosted elsewhere (GitHub? A blog?) and referred to here if it is continually being edited. MO is not really for such material, but for a focussed Q&A format $\endgroup$
    – David Roberts Mod
    Commented Dec 13, 2022 at 20:24
  • 1
    $\begingroup$ Links can be a solution, and while we all use links for convenience there is a general consensus in stackexchange "try not to use links"/"try have answer contain all their relevant information". I like the idea of a programatic muting a lot since it seems like an easy way for two groups of people to both be content. $\endgroup$
    – Sidharth Ghoshal
    Commented Dec 13, 2022 at 20:26
  • 1
    $\begingroup$ Also, just because something is answered doesn't mean the answer cannot be improved. You maybe dont want a FINAL edit. If the question asked "what is the list of superconductors above 100K" then an answerer might want to make that into a continuously updating list as new superconductors get discovered etc... in a math specific context this might be "how many primes satisfying obscure property X are there" and someone might answer only prime "P. but a year later "P, Q", and then "P, Q, R" .... as new results get discovered. $\endgroup$
    – Sidharth Ghoshal
    Commented Dec 13, 2022 at 20:28
  • $\begingroup$ Still neither of those apply to this case where the asker was updating over a much shorter span of time. I still think that a good thing happened when anixx made those 35 edits but it happened in a way which annoyed of bunch of people (because of that repeated refresh to the home page). So I want to enable users like anixx to do that without bothering people. $\endgroup$
    – Sidharth Ghoshal
    Commented Dec 13, 2022 at 20:29
  • 2
    $\begingroup$ If SE were to allow this, how would one prevent abuse of the system? And your example about "a year later..." seems to indicate sporadic updates. This is not really the case here, rather dozens of edits in the space of a coupleof months. And there's also a difference between a user with continual good contributions and respect in the community, and, I'm sorry to say, someone who has required extensive patience. $\endgroup$
    – David Roberts Mod
    Commented Dec 13, 2022 at 20:34
  • 1
    $\begingroup$ What would constitute abuse exactly here? Suppose we have the muting in place on updates to the answer. Then either 1. the answer has useful content and keeps getting updated without bothering anyone or 2. the answer doesn't have useful content and starts attracting downvotes (and if the user accumulates enough of this then they are booted off the site). $\endgroup$
    – Sidharth Ghoshal
    Commented Dec 13, 2022 at 20:37
  • 2
    $\begingroup$ This is the internet, I hope you can imagine for yourself what might be done on obscure pages that don't get periodic sunlight that the edit->front page process gives. This mechanism you want is not something no one has asked about before, and it's not an MO-level decision. $\endgroup$
    – David Roberts Mod
    Commented Dec 13, 2022 at 20:41
  • 1
    $\begingroup$ I can refine my proposal: if a page is obscure and getting one edit out of the bloom it'll go to the front page as we normally are used to. If a page is obscure and suddenly gets an influx of edits, for some period of time T it will stop getting bumped to the front page no matter how many edits are made to that answer (it already got one when the influx started). Once that period of time T passes it gets another "front-page-opportunity" from that answer. So we aren't preventing sunlight, we are just trying to reduce noise. You can call this a "refresh"-refractory-period. $\endgroup$
    – Sidharth Ghoshal
    Commented Dec 13, 2022 at 20:47
  • 2
    $\begingroup$ Somewhat similar suggestions seem to reappear regularly. Feature request by a former moderator: Minor edits, subject to review. And a similar feature request on Meta Stack Exchange: Allow non-bumping minor edits, but review them on /review. $\endgroup$
    – Martin Sleziak
    Commented Dec 13, 2022 at 21:16
  • 5
    $\begingroup$ I used to say, in the old days, "MathOverflow is not a blog" and I still believe that. $\endgroup$
    – Yemon Choi
    Commented Dec 14, 2022 at 2:04
  • 1
    $\begingroup$ @YemonChoi, re, I definitely agree that frequently updated answers are often a symptom of blog-giness, and, to the negligible extent to which my opinion matters, I do not support this particular proposal (besides which, it having been, as @‍MartinSleziak has indicated, so often proposed before, it seems unlikely that this will be the time it is implemented)—but the very specific use case here seems to be decidedly un-blog-gy, and perhaps more like software maintenance. $\endgroup$
    – LSpice
    Commented Dec 14, 2022 at 2:16
  • 1
    $\begingroup$ Sometimes I do this: I write an answer. I delete it immediately. Then I edit it (again and again, as much as I want). Finally, when it is ready, I undelete it. $\endgroup$
    – Gerald Edgar
    Commented Dec 14, 2022 at 10:51
  • 3
    $\begingroup$ @GeraldEdgar those edits will still bump the post to the frontpage (as per the FAQ on bumping on Meta SE). $\endgroup$
    – The Amplitwist
    Commented Dec 14, 2022 at 14:31
  • $\begingroup$ A compromise to @GeraldEdgar's solution, based on @‍TheAmplitwist's information—which I didn't know; thank you!—is to not save each edit. (I think SE uses local storage to preserve un-saved edits to a post; if I draft an answer but don't post it, then it's usually still there when I re-visit.) Or, of course, an easier approach would be not to post the answer at all, and only keep a local copy, which you can edit with abandon. Also relevant? $\endgroup$
    – LSpice
    Commented Dec 15, 2022 at 22:07

1 Answer 1

Reset to default
4
$\begingroup$

Let me try to mention a few things. Specifically, I want to concentrate on:

  • Some statistics about bumping/editing.
  • Already existing feature requests.
  • What users already can do to avoid excessive bumping.

Some statistics.

SEDE contains various information about Stack Exchange posts, including the PostHistory table, which contains some information about the past revisions. From this we can get number of revisions of the post. The bumps by the Community User count as revisions - most of the queries below count only edits (including the initial version and rollbacks).

Notice that several of the queries have various parameters that you can change. For some queries, you can chose whether to look at all posts or only answers/questions.

Other related feature requests.

Some suggestions that something like having edits which would not bump the post to the frontpage were posted in the past - with various proposals how to deal with them.

If you consider those suggestion useful, you might upvote them - at least to signal support, if SE considers implementing some of them at some future point.

Although it seems that the main concern in those discussions is edits on the posts made by other users.

Workarounds in the current situations.

Still, if somebody is running into the problems with the number of edits, they might try to limit it in various ways. If nothing else, one could try to fix multiple things at the same time - one way to do that would be to keep a copy of the post elsewhere, where edits are not disruptive; and only update the MO post after several edits on the other copy.

  • One could simply keep the post locally on their computer.
  • One could use some editor which supports the same (or at least similar) syntax as Stack Exchange. I am familiar with StackEdit - but probably there are other suitable options, too. See also: Is here something like blog service or just sand-box to store notes ?. And on Mathematics Meta: MathJax WYSIWYG Editor, MathJax: better way to prepare a Math.StackExchange question?.
  • On Mathematics Meta there is this post: Sandbox for drafts of long, complex posts. (I am not sure to which extent this can be considered MO analogoue: Formatting Sandbox.) This is probably only useful before posting the first version.
  • This is probably not optimal, but at least in theory one could post a question and then an answer. After deleting both of them, they could serve for them as a "personal sandbox". (It is not possible to edit a self-deleted question - but one can edit their deleted answers. If the question is deleted too, this won't disrupt the frontpage. However, I am not sure whether or not this usage of deleted posts is frowned upon.) I will explicitly mention that if a question is not deleted, editing a deleted answer still bumps the question. (I am mentioning this mainly because editing an answer while it is deleted was suggested in the comments.)

Another thing which one could use would be using saves. (This feature is relatively new.) One could keep here one list with posts which need updating. This feature allows to add a private note to each post in the list - which would be a suitable place where to note what exactly needs updating. And then simply edit several things at the same time.

$\endgroup$
1
  • $\begingroup$ Just to make sure to give credit where it's due (though I am sure that, to someone with your expertise, it's fairly obvious, I suspect many people are in my shoes, in the sense of finding it surprising and unexpected behaviour), I note that @TheAmplitwist also mentioned the fact that editing deleted answers bumps the question. $\endgroup$
    – LSpice
    Commented Dec 30, 2022 at 21:15

You must log in to answer this question.

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