14
$\begingroup$

I noticed tonight that a well-intentioned user has been doing a lot of minor edits of formatting of old questions with the effect of filling most of the active question list with old questions bumped by minor edits. Did we ever reach some sort of consensus as too how many minor edits a user should restrict themselves to in a single day? I know on the old meta, now known as tea, such issues were discussed.

$\endgroup$
  • 4
    $\begingroup$ In my opinion one should not do minor edits to more than 3-5 old posts in a 5 hour time period. Doing say 8 minor edits in a short burst can fill almost the entire active question list with old questions which means that new questions may be missed. $\endgroup$ – Benjamin Steinberg Aug 8 '13 at 4:02
  • 5
    $\begingroup$ I agree that bumping lots of posts to the front page is undesirable. Nevertheless, I would like to say that at least several of the edits in question corrected mathjax rendering problems. In fact, it was this recent series of edits which led me to start the thread meta.mathoverflow.net/questions/598/…. $\endgroup$ – Ricardo Andrade Aug 8 '13 at 4:45
  • 4
    $\begingroup$ Why does editing bump questions to the front page? Is there an official explanation? Why does an edit cause the post to be bumped, but a comment doesn't? Just to be clear, I'm not saying a post should or shouldn't be bumped when edited, I'd just like to see the reasoning behind it. $\endgroup$ – Michael Albanese Aug 8 '13 at 11:57
  • 1
    $\begingroup$ @MichaelAlbanese Of course, everything depends on the way you are viewing questions (whether you choose newest tab, active tab, ...) A list of things that cause bump is given at meta.SO: What can cause my question to be bumped? See also tag-wiki for (bump). $\endgroup$ – Martin Sleziak Aug 8 '13 at 12:13
  • 7
    $\begingroup$ It's there to stop malicious edits from going unnoticed. I'm not sure why it's still necessary now that low rep edits go through an approval process. $\endgroup$ – Noah Snyder Aug 8 '13 at 14:05
  • 4
    $\begingroup$ @NoahSnyder: I agree with the first sentence. But do not understand the second one. Before low-reo users could not edit others' posts at all and editing their own does not go through review. (So there seems no 'security-improvement' relative to before.) $\endgroup$ – user9072 Aug 8 '13 at 15:48
  • 2
    $\begingroup$ I would like to underline one point @MartinSleziak mentions in passing, namely, that 'active' is not the only tab (but precisely the one that shows "every" activity). When I do not have much time for visiting MO frequently, I browse newest where I get all (new) questions ordered in reverse chronological order (nothing whatsoever causes any bumps in that tab). Newest is a bit limiting, but 'hot' could be a reasonable compromise (not everything that bumps in 'active' bumps in 'hot', eg pure tag-edits do not show up in 'hot'). $\endgroup$ – user9072 Aug 8 '13 at 16:10
  • 7
    $\begingroup$ @NoahSnyder: The approval process would be enough if all reviewers actually looked at what they approve. But experience shows that some of them don't, as witnessed in this discussion on meta.SO. People manage to approve suggested edits such as this one on other sites. $\endgroup$ – Martin Aug 8 '13 at 16:17
  • 3
    $\begingroup$ @quid, but active is very convenient for seeing new answers. If you answer a question and it immediately gets bumped off the active page then few people will see your answer. $\endgroup$ – Benjamin Steinberg Aug 8 '13 at 16:19
  • $\begingroup$ @BenjaminSteinberg: yes, I partly agree. Still, sometimes one can get away with the impression that some (not you specifically) do not really realize there are other tabs too, and since you explcitly mentioned the risk of missing a question I mentioned it. However, I am also not sure the effect is as large as you describe. Namely, I think it is an oversimplifcation to assume that there is an absolute or even very strong correlation with things getting looked at an their precise position in the active queue. Sure I an I assume many/most others start looking at the top of it, but then [cont] $\endgroup$ – user9072 Aug 8 '13 at 16:47
  • 1
    $\begingroup$ A bit off-topic here but just as a vague idea, perhaps to get realized a feature-request for an additional tab (say, containing only new questions and answers), or a replacement of another one (say 'hot') by this, would have more chance of success than the 'minor edits should not bump' request. $\endgroup$ – user9072 Aug 8 '13 at 17:03
  • 1
    $\begingroup$ @quid: I would upvote a feature request for a new tab like that. Unfortunately, such a feature is unlikely to ever see the light of day, judging by how most feature requests are ignored... $\endgroup$ – Ricardo Andrade Aug 8 '13 at 17:53
  • 2
    $\begingroup$ How about a "minor edit" flag? $\endgroup$ – Alexey Muranov Aug 12 '13 at 13:25
  • 3
    $\begingroup$ @RicardoAndrade: I would love to see a feature request that asks for feature requests to be addressed! $\endgroup$ – François G. Dorais Sep 6 '13 at 13:44
  • 3
    $\begingroup$ I think filling the active questions with lots of edits is particularly bothersome for those who mostly us MO on mobile devices because only a handful of questions are visible without scrolling in mobile mode. $\endgroup$ – Benjamin Steinberg Sep 12 '13 at 13:28
9
$\begingroup$

The OP wrote: I know on the old meta, now known as tea, such issues were discussed. I am collecting here some relevant threads which I was able to locate. I am making this CW - feel free to add more links, if you consider them relevant.

  • Editing old answers by other users; François G. Dorais said there: "It's fine and encouraged to improve old questions and answers. However, keep in mind that it's mildly annoying to have many old questions bumped up to the front page at once, 3 edits per day is acceptable."
  • Must we raise the dead for umlaut's sake; Gerry Myerson's advice: "Re-tag two or three questions a day until they're all done."
$\endgroup$
  • 1
    $\begingroup$ thanks for finding these. $\endgroup$ – Benjamin Steinberg Aug 8 '13 at 14:08
  • $\begingroup$ Just to confirm, the first is the last official (ie, moderator) mention of this I remember (and I am pretty sure it is actually the last) so it might qualify as current "official guideline." $\endgroup$ – user9072 Aug 8 '13 at 16:03
  • $\begingroup$ I agree with the "two or three a day" limit. For ALL edits of stale questions, major or minor. $\endgroup$ – Gerald Edgar Sep 6 '13 at 13:15
  • $\begingroup$ It's odd to be reminded of the verse / I used to try; but now it's even worse $\endgroup$ – Yemon Choi Sep 12 '13 at 13:52
12
$\begingroup$

I believe that I am the user that you are referring to. I apologise to anyone I have inconvenienced.

The majority of the edits that I have done over this short period of time were fixing very similar errors which caused the output to split over several lines. I think I got carried away because I wanted to fix all of them (there aren't that many to fix). However, I did not think about the effect I was having on the front page. I will take a break from these edits for the rest of the day, but (if it's OK with everyone) I will continue to do one or two a day so that all instances of this error are rectified within a week or so.

$\endgroup$
  • 1
    $\begingroup$ Dear @Michael: I would greatly appreciate it if you could describe those mathjax problems in the thread meta.mathoverflow.net/questions/598/…. $\endgroup$ – Ricardo Andrade Aug 8 '13 at 5:41
  • 14
    $\begingroup$ Michael, I would like to thank you for your efforts to edit and improve questions and answers here on MO; it is appreciated by many of us, and I hope you continue to do this. $\endgroup$ – Joel David Hamkins Aug 8 '13 at 5:57
  • 7
    $\begingroup$ These are needed edits. It would be nice if they could be done automatically. Since that does not seem to be the case, this seems the best solution we currently have. $\endgroup$ – Andrés E. Caicedo Aug 8 '13 at 6:10
  • 4
    $\begingroup$ I have no problems with the edits. I just recommend in smaller bursts, maybe 3 at a time. $\endgroup$ – Benjamin Steinberg Aug 8 '13 at 11:04
  • $\begingroup$ @RicardoAndrade: Will do. $\endgroup$ – Michael Albanese Aug 8 '13 at 11:06
  • 1
    $\begingroup$ @JoelDavidHamkins: No worries. I will continue to edit posts but I will try to space them out. $\endgroup$ – Michael Albanese Aug 8 '13 at 11:14
  • $\begingroup$ @BenjaminSteinberg: I agree, I will be more mindful in future. $\endgroup$ – Michael Albanese Aug 8 '13 at 11:30
10
$\begingroup$

Today, I have spotted Pietro Majer bumping close to 20 (and counting...) of his answers to the front page with minor edits. (Sorry to be so blunt in naming a respected user, but I guess I'm a little annoyed.)

$\endgroup$
  • $\begingroup$ Perhaps a comment on one of his answers might get quicker results? Hey, @Pietro! Slow down please! $\endgroup$ – The Masked Avenger Sep 5 '13 at 16:30
  • 3
    $\begingroup$ @TheMaskedAvenger I did that. No response. $\endgroup$ – Todd Trimble Sep 5 '13 at 16:32
  • $\begingroup$ @ToddTrimble Just seen this (this follows on from some minor but justified edits yesterday). I'm not keen on the extent of today's bumping either $\endgroup$ – Yemon Choi Sep 5 '13 at 16:36
  • 2
    $\begingroup$ After analysing a few of Pietro's edits, it appears he is mostly replacing double backslashes \\ in mathjax: this seems to now be non-functioning, yet harmless, code. In an attempt to prevent this from happening again in the future, I bumped my request to automatically fix these minor mathjax issues to the top, and turned it into a feature request. $\endgroup$ – Ricardo Andrade Sep 5 '13 at 17:30
  • $\begingroup$ In the meantime, I also left a comment in one of Pietro's recently edited answers, as Todd Trimble and Yemon Choi did before me. As it stands now, the front page consists mostly (~35/50) of posts with answers which have been edited by Pietro less than two hours ago. $\endgroup$ – Ricardo Andrade Sep 5 '13 at 17:49
  • $\begingroup$ @RicardoAndrade: Thank you! That's a good request. $\endgroup$ – Todd Trimble Sep 5 '13 at 18:13
  • 1
    $\begingroup$ @Todd Trimble: Let us hope that the developers pay attention to the request... By the way, for the record, there has been no new edit by Pietro Majer in the last 25 minutes. $\endgroup$ – Ricardo Andrade Sep 5 '13 at 18:17
  • $\begingroup$ @RicardoAndrade Heh. It was good while it lasted. (Actually, I'm rather surprised, as I even went so far to flag one of the edited answers to get moderator attention, and figured the hiatus in activity was due to a discreet request from a moderator). $\endgroup$ – Todd Trimble Sep 5 '13 at 18:56
  • 6
    $\begingroup$ Hi everybody, I noticed your messages and I understand now the terrible effect of my recent minor edits (I always keep the "newest" list and noticed nothing). Anyway, my excuses for the trouble! This afternoon I had some free time and thought to fix the format of old posts of mine once and for all (after the recent upgrade of the system, the // does not produces a space anymore, but breaks the lines, which is somehow annoying). Cheers, P. $\endgroup$ – Pietro Majer Sep 5 '13 at 19:43
  • $\begingroup$ @PietroMajer: many thanks for your response! I understand your own annoyance. Hopefully Ricardo's request can be implemented. $\endgroup$ – Todd Trimble Sep 5 '13 at 19:55
  • 5
    $\begingroup$ I said above that \\ (double backslash followed by a space) is harmless. However, that statement was not correct. It actually creates a new line when mathjax is rendered using CSS or using SVG. Interestingly, when using mathml to render mathjax in at least two browsers, it appears that \\ is simply ignored and does not create a new line. $\endgroup$ – Ricardo Andrade Sep 5 '13 at 20:45
9
$\begingroup$

It's happening again, more than 15 minor edits by one user. Clearly it is just not well known how much this affects the "Active" view. The view by newest Questions is unaffected. I have no suggestion, aside from revisiting "Silent” Edits for MO?"

$\endgroup$
  • 1
    $\begingroup$ Maybe ońe should contact the editor. These edits are really minor and the questions old. $\endgroup$ – Benjamin Steinberg Sep 12 '13 at 13:06
  • 4
    $\begingroup$ @BenjaminSteinberg: The editor is not the key problem, it's the reviewers who approved the edits who did them all at once. $\endgroup$ – François G. Dorais Sep 12 '13 at 13:14
  • 1
    $\begingroup$ @François G. Dorais: That’s true, but OTOH notifying one editor of the problem is feasible, whereas notifying all potential reviewers is not. $\endgroup$ – Emil Jeřábek Sep 12 '13 at 13:17
  • 2
    $\begingroup$ @FrançoisG.Dorais, it is both but I agree with Emil. I normally reject all minor edits, especially on old questions but I suspect I am in the minority and in any event I almost never review because in the mobile version of the site the Review section is not so accessible. $\endgroup$ – Benjamin Steinberg Sep 12 '13 at 13:25
  • 2
    $\begingroup$ @EmilJeřábek: I can (and will) notify the reviewers in question. I don't think it's their fault though since the reviewing system encourages dealing with these in batches. $\endgroup$ – François G. Dorais Sep 12 '13 at 13:28
  • 2
    $\begingroup$ @FrançoisG.Dorais what course of action would you propose for a reviewer in such a situation. I can approve/reject/skip. If I reject the edit, it is "gone" I think (except somebody else approves it of course) which seems highly inefficient. Now, if I skip, somboy else will get it possibly not being aware of the earlier edits or in any case in the same situation. Both to reject an edit because of timing (and thus 'destroying' the work that went into it) as well as some mass-skiping to spread it out seems unreasonable or infeasible. $\endgroup$ – user9072 Sep 12 '13 at 13:54
  • 2
    $\begingroup$ I have just rejected a POINTLESS bit of TeXifying, and think those of us who find such edits should reject, contrary to @quid's remark $\endgroup$ – Yemon Choi Sep 12 '13 at 13:57
  • 1
    $\begingroup$ @quid: Only a handful of users regularly look at suggested edits and most deal with them in batches. Raising awareness won't eliminate the problem but it will have some effect. $\endgroup$ – François G. Dorais Sep 12 '13 at 13:58
  • 2
    $\begingroup$ I should emphasise that my distaste for excessive editing here arises from my past experience doing some freelance copy-editing, which has led me to (a) believe in house style (b) wish sleeping dogs were left to rest unless waking them serves a purpose beyond making the editor feel pleased/busy/smug $\endgroup$ – Yemon Choi Sep 12 '13 at 14:00
  • 1
    $\begingroup$ @YemonChoi if you think the edit is "POINTLESS" it is of course your right or perhaps even duty to reject it. What I said or at least meant to say is that I do not think it is a good idea to reject an edit you/I/one would otherwise approve only because of timing. $\endgroup$ – user9072 Sep 12 '13 at 14:05
  • 1
    $\begingroup$ @FrançoisG.Dorais not at all sure about this. I am third by number of reviews so can be considered very active there. Hardly ever do I come across there being more than 2 pending suggestions. More generally, most of the time all suggestions are dealt with very or quite close in time they are suggested. Some are more one the site and/or quicker to react than others, but in principle I think not few look at these suggstions, also as the pending suggestions are made so very visible (as opposed to other review types). $\endgroup$ – user9072 Sep 12 '13 at 14:12
  • 1
    $\begingroup$ @quid: Indeed, you almost never review more than 2 edits in a row. Other users do long batches of 5 or more in a row. Those are the reviewers that need to be aware of this issue. $\endgroup$ – François G. Dorais Sep 12 '13 at 14:19
  • 4
    $\begingroup$ @KarlSchwede. I think it is a real problem. At the moment the oldest post on the front page is 6 hours old and on a regular day the oldest would be maybe twice that. If you're asking a technical question that only one or two users on the site know how to answer, then the chance that one of these people sees your question drops drastically if posts get bumped off the front page this quickly. $\endgroup$ – Dan Petersen Sep 12 '13 at 14:22
  • 1
    $\begingroup$ @DanPetersen I understand, I just tend to look only at tags I am interested in, via the "newest" tab. In which case this sort of thing never shows up. I realize that other people display things differently, but I'm not quite sure what the correct thing to do is (see quid's comments). $\endgroup$ – Karl Schwede Sep 12 '13 at 14:27
  • 2
    $\begingroup$ @FrançoisG.Dorais I do not want to insist too much on this, but I just roughly checked the history, ignoring the last hours, the first time I found a string of 5 was on August 23rd! There was none of size 4 until that time, and a handful of 3-strings. It simply hardly ever happens that anybody would even have the chance to make such a long string as typically there simply is no sizeable batch. $\endgroup$ – user9072 Sep 12 '13 at 14:43
8
$\begingroup$

Some time ago, still unaware of the mechanism of the active tab, I had the idea of correcting the wrong spelling of the name "Hölder", too often written "Holder" by laziness or ignorance, in every occurrence in MO posts. As a result, we had a huge wave of Hölder-related questions over all MO users, a fact I was sincerely sorry about (my excuses again). However, I completely forgot the incident and unfortunately I happened to do the same some days ago, this time in order to eliminate the annoying double backslash. I am now sure that I will never do that again. Nevertheless I think the MO archive of questions and answers is a very rich and useful source, so that we should find a way to allow and even promote improvements of older posts, especially about making corrections and adding new links and references. Certainly, a good compromise is to introduce a bound on the number of minor edits. Yet some edits, like the above mentioned, would be done better if all at a time, once and for all. So my modest proposal is the following: Allow free edits on some special days. How often, it has to be decided. So everybody will be prepared to see, say, on the first Sunday of every month, the active tab turned into a total mess, wildly shaken all day long.

$\endgroup$
  • 3
    $\begingroup$ Dear @Pietro Majer: Your proposal may warrant creating a new thread for its discussion. $\endgroup$ – Ricardo Andrade Sep 7 '13 at 0:11

You must log in to answer this question.

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