260

On Stack Overflow, there was a question with an incorrectly formatted list, which lacked a newline before it and thus came out like this: 1. Hello 2. Something else 3. Goodbye

I changed it to the right formatting by inserting a newline:

  1. Hello
  2. Something else
  3. Goodbye

And I got the "Edits must be at least 6 characters" error message. Is there an OK way to trick this rule?

  • 56
    Just remove "Hello" and "Thanks" in top and bottom of the post :) – BalusC Mar 3 '11 at 14:49
  • Related: meta.stackexchange.com/questions/82534/… – BalusC May 5 '11 at 21:34
  • 168
    This is annoying for source code changes, where only a ; is missing … – Martin Ueding Mar 6 '12 at 15:41
  • 91
    It makes me angry, too. If I see a simple syntax error, like a misplaced comma, I then have to start adding noise to the post just to make the edit stick. Even worse when a poster doesn't know how to use indenting and then the blank spaces don't even count towards the 6 characters. I like SO but not this rule. – fritzfromlondon May 7 '12 at 21:56
  • @fritzfromlondon I think the accepted answer has the solution: there is always something else to do even if it is not your primary objective when pushing the edit button. You can change the text a little bit to make it more readable or undertsandable. – vbence May 8 '12 at 11:57
  • 38
    Rules with magic numbers (like 6 in this case) are always bad rules. – dokaspar Aug 6 '12 at 7:42
  • @Dominik SO likes to have constraints on pretty much every variable not to get out of control (like daily rep cap or vote cap). It's not necessarily a bad thing. – vbence Aug 6 '12 at 9:57
  • 56
    @vbence: No, there isn't always something else to do. Sometimes there is, often there isn't. Preventing people from making important edits because they can't find any additional inconsequential edits to make is stupid. – endolith Oct 10 '12 at 21:39
  • 2
    vbence, please, for the love of god (or the hate of god, or the apathy towards god, whichever you like best), accept the better (next-highest) answer. – codetaku Jul 28 '14 at 17:59
  • 7
    I found an answer in which both the terms "CPU" and "GPU" were used, since it was a question about both. In the otherwise good answer, the author used the term "GPU" instead of "CPU". While this is only a 1 character error, it drastically changes the meaning. I'm not sure how to get around this, so it's left unedited... – orrymr Mar 13 '15 at 13:10
  • 3
    It's not just formatting or typos, either. A lot of the time- for example, in shell scripts- the entire thing is wrong if it forgets to put double quotes around a variable or lacks an & at the end of a line. – Parthian Shot Aug 13 '15 at 22:35
  • 1
    @dokaspar: Rules like you cannot drink booze under age of 18/21? – Vajk Hermecz Jan 5 '16 at 12:36
  • 1
    @VajkHermecz, that rule seems to be a good example: it is clear that there is no magical transition that happens upon turning 18 in some countries, or 21 in others. A (logically) better rule would be some sort of competency-oriented test, but of course that would be an impracticable nightmare to implement, so we settle for a bad test as better than no test. – LSpice Feb 12 '16 at 19:46
  • 1
    there is a way to trick it. – John Militer May 9 '17 at 2:50
  • 3
    The 6 character rule discourages people from making valid/needed changes--and actually starts working against quality--because for posts that have already been reviewed by hundreds or thousands of people already, there actually aren't too many ways to make these posts better....it's usually just some some non-major (but still important) thing that needs changing and we don't want to add noise just to make that change because it would decrease the quality of the post. So the minor mistakes remain. Why not make the number 1 instead of 6? – jbobbins Nov 3 '17 at 15:59

10 Answers 10

53

Sure! Instead of tricking the rule, just look for other changes that can be made. Usually, in a post that forgets that kind of formatting, there's often at least one other error to be found. A miscapitalized letter, an extra space, or often just another formatting error of a different sort.

Remember, when you suggest an edit, it requires multiple other people to look at it and approve it. The character limit is to prevent people from wasting time by looking at exceptionally minor edits. So, don't limit yourself to just a tiny edit: try to see if you can improve the post to a possible state of perfection. If you hit all errors on a post, then no one else will even need to edit it.

Once you hit 2k reputation, and thus your edits don't need to go through the approval process, you can make those tiny changes without the limit in the way.

  • 5
    Are you sure of "no limit over 2k"? I seem to be hitting the limit, even at 16k - but as it does prompt me to make other changes, it makes me realize that any edit should improve the post considerably, therefore I do more editing. – Piskvor Mar 3 '11 at 15:04
  • 5
    @Piskvor - I can confirm that the 6 character limit doesn't apply to 100K users. ;-) ok, technically, 133K users, but I doubt anyone would pick 133K as an arbitrary cutoff, maybe 128K, but not 133K. – tvanfosson Mar 3 '11 at 15:15
  • 6
    I beg to disagree just because code snippets need spaces in front. So space only changes are not uncommon. In fact, you got it the wrong way around: When all you do is add spaces then then you can't do any evil with it. No 2k users needed. BTW: not knowing the background all i did make a useless 6 character change and a rant in the explanation. Now there will be some 2k users having fun now. – Martin Apr 27 '11 at 16:52
  • 60
    But what if there aren't any other changes to be made? The people who make up these rules don't seem to think them through before implementing them. – endolith May 5 '11 at 21:18
  • 82
    As a new user I can say this limit is rather annoying. In StackOverflow I saw a code snippet where someone had an extra e in their code. All I wanted to do was remove the extra e so the code would compile. It would not let me change a reference of Neested to Nested to prevent a compile error. – JSWork May 19 '11 at 19:45
  • 7
    @Grace: How do we "let a 2k user handle it" if they aren't aware of the problem? I thought the whole point of this limit was to prevent bothering 2k users who have to review every edit anyway. – endolith May 27 '11 at 21:22
  • 7
    Small changes can make the difference between correct and incorrect code. For example, this XAML here needs quotes to work: stackoverflow.com/questions/2948290/… I didn't see anything else to improve, but it would be a convenience to other people copying and pasting the code to try it out, if the code were correct. – Edward Brey Aug 19 '11 at 15:14
  • 87
    try to see if you can improve the post to a possible state of perfection That is the dumbest thing I've read all day. If you edit a post, you should ONLY change specific and clear problems. such as un-highlighted code, or spelling the word "string" as "dtring". When you try go beyond discrete problem areas, you risk changing the question to something that the OP didn't intend. – Sam I am Mar 21 '12 at 19:57
  • 4
    would be interesting to know how do you submit "improved" edit suggestion for this code: stackoverflow.com/a/8722074/155687 - there is ONLY ONE character to delete (mismatch in regex and error message) <asp:RegularExpressionValidator runat="server" ID="valInput" ControlToValidate="txtInput" ValidationExpression="^[\s\S]{0,100}$" ErrorMessage="Please enter a maximum of 1000 characters" Display="Dynamic">*</asp:RegularExpressionValidator> – Vladislav Aug 30 '12 at 13:17
  • 13
    I would argue against the "state of perfection." Perfection doesn't matter. Editing to that extent has diminishing returns. – Austin Henley Dec 27 '12 at 19:54
  • 12
    At first, your answer is dumb. It just proves that dumb requirements are supported by dumb arguments. Secondly, I could understand if you say that this filter stops minor edits. But, you recommend doing extra useless exits to pass the filter. This is opposite of saving everybodies time! – Val Jul 2 '13 at 15:38
  • 7
    @OP: "Instead of tricking the rule, just look for other changes that can be made [...] there's often at least one other error to be found". -> That's simply not true. – Jocelyn Jul 24 '13 at 13:27
  • 10
    sigh This is painful when someone whose first language is probably not English has given a good answer - except for confusing a pair of English homonymns, creating nonsense. "Exits" and "exists" in this case. Even with my proofreader's eagle eye, I only found one other similar mistake (its for it's IIRC) and had to change a perfectly valid word to a synonymn to increase my character count. Even with my explanation in the edit reason, reviewers are going to see the silly change first - the important typo was in the last line of the answer - and probably disallow the edit. sigh – Arlie Stephens Aug 25 '13 at 19:01
  • 5
    I'm surprised that the comments on this answer are almost exclusively negative. I think the motivation is that setting a minimum does more good by preventing a flow of extremely minor / pedantic edits (especially from new users), than it does evil by bothering users (who were already willing to click and punch a few buttons to edit) to find 1 or 2 more things to edit. I get that one character makes all the difference in code, but most edits are not of that nature, and even among the ones that are, we can almost always find something else to edit, in less than 15 seconds, and – Andrew Cheong Dec 14 '13 at 9:48
  • 10
    -1 I also dislike the "rules are smarter than reality" attitude of this rule. At the very least, there should be "I'm sure there is nothing else to fix" override button. – Aaron Digulla Jan 29 '14 at 10:28
174

To circumvent this counterproductive rule without confusing the recipient of your edits, just add a &nbsp; or HTML comment <!----> at the end of a line, where it won't make any difference to the formatting.

If Stack Exchange adds a rule to ban those (because they really prefer that you don't fix typos and mistakes), then you'll have to be creative with other non-printing or whitespace elements.

The How to Edit box next to the edit window encourages exactly these kinds of changes:

How to Edit

► fix grammatical or spelling errors

► correct minor mistakes

So it's pretty hypocritical for the site to reject your helpful contributions after explicitly asking for them.

  • 1
    Another thing you could add: meta.stackexchange.com/a/91535/130885 – endolith Nov 21 '12 at 18:47
  • 27
    Nice hack. It's a completely stupid rule when you want to change a code error such as "dict*" to "*dict" – martinjbaker Feb 14 '13 at 14:08
  • 2
    ... or "dic_" to "dict" – user208043 Jun 20 '13 at 3:00
  • 19
    This needs to be the accepted answer for so many reasons. – codetaku Jul 28 '14 at 17:55
  • Probably I missed something, I didn't understand how to use &nbsp; but this empty space is OK, I copy it by holding SHIFT and pressing arrows fileformat.info/info/unicode/char/200B/browsertest.htm: ->​<-. BTW this is not only SO rule... I've seen multiple forums and communities with this, I really wish a SO admin to comment on this: WHY? – ivkremer May 15 '15 at 11:50
  • 1
    @Kremchik You just type &nbsp; somewhere at the end of a line. – endolith Jun 16 '15 at 4:20
  • 1
    It's even better to add an HTML comment (<!---->) because it doesn't display at all. – Donald Duck Aug 17 '17 at 8:05
  • Another example of nonsensical edits to meet the quota for a single character change: dsp.stackexchange.com/posts/472/revisions – endolith May 4 '18 at 23:50
35

It is incorrect for a 6 character limitation to be imposed. When you edit at top of the page there is:

We welcome all constructive edits, but please make them substantial. Avoid trivial, tiny one-letter edits unless absolutely necessary.

So what to do when absolutely necessary? That limitation needs to be removed and normal approval process take place or the above text needs to be changed with other suggestions.

P.S. I liked the answer to add non-printable characters to post. I didn't think they were counted.

  • 5
    But the point is that SO won't LET you do a 1 letter edit without resorting to these silly hacks. And it isn't "trivial" if there's a typo in a code example that simply won't work because of it. – martinjbaker Nov 20 '13 at 18:34
  • 29
    A tiny edit is not necessarily a trivial one (if you ask a compiler). – Anders Lindén Dec 14 '13 at 8:14
  • 4
    I wanted to fix a typo of were to where, but I couldn't, because 6char. Thanks, Stack Overflow. – Quolonel Questions Dec 11 '16 at 14:16
30

I disagree with the premise that small edits are a bother to the reviewers. I only recently hit the threshold (for having edits applied directly and for reviewing others) and I think the smaller the edit is, the easier it is to approve. If I see a small typo being fixed, code formatted, or a sentence properly capitalized, I can see right away what it is and approve it.

The bigger the edit is, the longer it takes to check.

I say the 6 character limit needs to be abandoned.

  • It is easier and quicker to review one edit rather than several - you have to read the whole question/answer to make sure that it is a reasonable edit - if you are not doing this then you are not reviewing – Mark Mar 1 '14 at 20:01
  • 6
    @Mark But I'm not suggesting to spread out one big edit among several small ones, I'm arguing that sometimes a small edit is all that's needed. And if I see that the only difference is a typo being fixed, I don't have to read the entire post to judge the edit. – SQB Mar 1 '14 at 21:45
  • The issue raised by others is if there is an error in formatting then there is likely to be other errors - so a small edit is likely to be a too minor edit – Mark Mar 1 '14 at 22:17
  • 5
    @Mark Yes, but that's not always the case. Sometimes it's just the one typo that needs to be fixed to have the answer make sense. – SQB Mar 1 '14 at 22:21
  • 7
    I agree with your premise, @SQB, and let me say also that the two rules "Edits must be at least 6 characters long" and "Above all, respect the original poster" could potentially be at odds with each other. Sometimes scouring the post for multiple mistakes or making some picky formatting change just to rise above 6 characters can be disrespectful of the author, unless it actually does improve something about the post; and likewise the correction of a "minor" error (as measured by the number of characters) can be much more respectful of the original poster. Just a thought. – Brian J. Fink Mar 3 '14 at 0:49
  • I don't think that changing a single "i" to "I" is worth 2 reputation points. This alone is reason enough to leave the limit in place. – ShaWiz Jun 24 '14 at 7:04
  • 5
    @ShadowWizard Then the system shouldn't award reputation points for single-character edits. You're conflating 2 different things. Small edits should be accepted regardless of whether they give rep or not, since they're beneficial and reduce reader confusion. – endolith Dec 5 '14 at 20:41
  • @endolith there's already such a feature request, but the thing to consider here is the required development time vs. how important this really is. With thousands of users actively editing and great majority of posts needing more than 6 characters edits, the team better spend their time on developing better things. – ShaWiz Dec 5 '14 at 20:47
  • 1
    @ShadowWizard "required development time vs. how important this really is" Accuracy is important. – endolith Jul 11 '17 at 15:21
  • 2
    @Mark In response to: "if there is an error in formatting then there is likely to be other errors", I don't think that's logically justifiable when a post has been seen by hundreds of people and edited by several. I found a post upvoted by 338 people. It was very well written and had probably been edited many times. But it still had only one issue I could see: the word "and" was used where it should have been "are". I had to read the phrase a couple times to know the intended meaning--I don't think that's "too minor" of an edit when probably 1000s of others are having to do the same thing. – jbobbins Nov 3 '17 at 16:29
22

Some people feel there should be an exception to this rule for single-character errors in code. (There is a general consensus that single-character errors in sentences are highly unlikely to matter.) My approach to this is:

If it's not your question, and you see a one-character typo, your ANSWER is "hey, OP, you have a one character typo: you have [line paste] where you should have [corrected line].

If it's your question, you update it and add "Update: Sorry, when sanitizing this code I made a one-character typo: the code that causes the problem did not have [error] where it should be [correction] and I've corrected my sanitized version in this question." Now the 400+ answerers who are telling you about the typo can delete their answers.

Either way, you avoid the single-character edit.

  • 18
    If it's your own question, then there is no 6 character limit. – Grace Note Jun 9 '11 at 18:06
  • @Grace - Thanks. I still feel people who edit typos out of the code in their questions should, if there are answers or comments referring to the typo, add explanatory text identifying the typo as a red herring. – Kate Gregory Jun 9 '11 at 20:18
  • 2
    @KateGregory: That's why you can flag comments as "obsolete" and have them removed. One of the big things that makes Stack Exchange better than forums is that you don't need to wade through pages of discussion to get the information you need. Users have a discussion about some point in an answer, the outcome of the discussion is edited back into the answer, and then the obsolete discussion is deleted. Distill out the important content and delete the chaff. – endolith Nov 23 '12 at 2:22
  • But then there are the idiots who get upvotes for /answers/, rather than comments, that achieve nothing other than pointing out the typo. How do we mark those as obsolete? – codetaku Jul 28 '14 at 17:57
  • 2
    @codetaku: Can you show an example? Flag them as "not an answer". – endolith Jan 29 '15 at 18:52
  • 2
    "there is a general consensus that single-character errors in sentences are highly unlikely to matter", have you got a source for that claim on consensus? I've just ended up here because someone's answer included a file path with a missing backslash somewhere in the middle of the long path that took me time to figure out the right place. And so it will for the next person. – Alan Macdonald Jun 3 '16 at 10:06
  • 1
    a file path is not a sentence. It's more like code, @AlanMacdonald. Other than typing now/not in "this is now possible" words in sentences tend to be pretty typo proof. – Kate Gregory Jun 3 '16 at 12:01
  • 1
    @KateGregory in my rage at the 6 character edit rule I never read what you were saying properly. I understand now thanks. – Alan Macdonald Jun 3 '16 at 12:12
14

I encountered the same problem, where the person that answered the question did a great job with the answer, but the links were off. So I had to edit just 3 characters.

It was an old post, I believe no one ever fixed it because of the 6 characters limitation.

It seems rather counterproductive that in order to fix a small problem you have to look around to be able to add valuable information.

What if you are like me? I found that question because I was learning about the subject. How could I add valuable information to an answer that was already correct except for the links being off? (btw just 2 links were off, so the answer had enough information to be understood).

I understand that there are people that review the corrections, and minor changes, like correcting a typo that is easy to understand would be hard to review, but I don't know, maybe something should be done for special cases like the one I had with the links.

  • 6
    I suppose the passive-aggressive approach would be to flag the post for ♦ moderator attention and describe the needed correction in the flag description. Overkill, sure, but at least the mods can make the edit. Not that I'd seriously advocate actually doing this. – Ilmari Karonen Dec 27 '12 at 20:39
  • 1
    More practically, you could describe the problem in a comment. If the author of the post is still around, they'll be notified about it, and in any case the comment will at least inform readers about the issue. – Ilmari Karonen Dec 27 '12 at 20:40
  • 7
    Well we could do all that and more. I looked for an additional link in my case. But It is an overkill! And it kinda goes against the website goal: "Stack Overflow is as frictionless and painless to use as we could make it." – Dzyann Dec 28 '12 at 12:41
  • 4
    Similar thing happened to me: I had a problem, found an old Q&A, followed a link to "more info", corrected the link as it didn't work, and ... had my edit rejected as being insignificant! That rejection was a greater time-waster (and insulter) than if it would have been accepted. Shame on the petty, hypocritical pedants at SO! – Martin F Dec 24 '13 at 22:49
13

The limit is stupid. Sometimes there are simple typos, easy to fix - but only one character. Post would improve. And nobody can tell me reviewers are overloaded because of this, because in SE-sites I can access the review-queue I have to submit much much less time with approving simple edits like fixing typos than with complicated edits. A fixed typo is easy to see.

Even worse, sometimes I find simple edits that fix something - and then do something with formatting or similar that makes the post worse. I assume that was done to reach the 6 characters.

Based on my experience as reviewer in some sites and suggesting edits in others I would say: make the edit and if it doesn't reach 6 characters add following to the end of the post:

<!-- character-limit: please remove -->

Reviewers have an 'Approve and edit'-button, so it is still simple enough to work with and it is easy to understand what the edit is for.

7

I have used the &nbsp; trick a few times, but I find myself adding the backticks: ` for code formatting around missed inline formats.

By highlighting these it only takes a few of them to help with the char count, but it also emphasizes often missed inline code examples.

NOTE: This is not a hack...
Rather, this is a way to add to char count by observing often overlooked formatting techniques.

  • 1
    If the downvotes here are to indicate "please don't abuse code formatting", please note that that's not what the author seems to be suggesting. If I understand correctly, he's merely stating that authors tend to occasionally forget to properly format their in-line code. In which case formatting it would be appropriate. Of course it isn't a general highlighting tool. – Bart Feb 28 '14 at 21:06
  • 1
    Thank you @Bart for clarifying.. I will update the post as such. i do not mean to "hack" the system, just add another suggestion for ways to add to character count that is often overlooked in the fury of typing the question first time through. – Phlume Feb 28 '14 at 21:25
  • 3
    It's just that we have rather a lot of formatting abuse @Phlume, where users highlight every other word with code formatting. So I can see how users might misinterpret your suggestion. If it's not code, don't format it as such. If it is, then it's a valid suggestion. – Bart Feb 28 '14 at 21:28
  • @Bart Agreed. Thank you for the help clarifying for the users.. not perhaps a reversal could be in order from the DV peeps?? – Phlume Feb 28 '14 at 21:40
7

This may be bending the rules, but here's a little trick I've found. You can edit the misspelled word, then add an empty HTML comment tag afterwards. For example:

The sum of it's effects

could be changed to:

The sum of its <!-- --> effects

which displays as:

The sum of its effects

It gets the job done for an edit of only one character. Then when the edit has been approved, you can reedit to remove the extra characters, if you want. Or it may be left in place, to be removed at the next legitimate edit.

But use this sparingly; if everyone used it just to circumvent the rule, it would eventually be banned. So use it only when you can change nothing else!

Of course, you don't have to put the comment tag close to the edit; it can actually be placed anywhere in the post.

  • @Bart you're the only commenter, so I'll ask you: should I delete the SE suggestion, or is my post completely irrelevant to the question? – Brian J. Fink Feb 28 '14 at 21:46
  • If you're referring to the downvotes, I can only assume they have been cast in disagreement with your work-around. If you feel it's a valid and acceptable practice, you can leave it in place. But the community doesn't seem to think it is. – Bart Feb 28 '14 at 21:49
  • One comment about my comment above: it refers to a suggestion to SE which is no longer in this post. I left it in place because it would be even more confusing if I didn't leave some context to @bart's comment. – Brian J. Fink Mar 3 '14 at 0:31
  • My trick is replacing any dot you want with a italic dot: <em>.</em> – Ooker Mar 16 '15 at 18:29
-3

You can use $\!$ to bypass the character requirement. Just add more \! as needed.

Note: This only works on sites with LaTeX enabled.

  • 4
    This seems to be strictly inferior to the already-given workarounds that don't require MathJax to be enabled. – Nathan Tuggy Jan 28 '17 at 22:46
  • 1
    &nbsp; &nbsp; &nbsp; – suomynonA Jan 29 '17 at 16:44
  • @NathanTuggy As you can see, the &nbsp; trick does not work in comments – suomynonA Jan 29 '17 at 19:33
  • 4
    But you can't edit others' comments, and as far as I know you can edit your own comments without character restrictions as long as you're within the 5-minute window, so this isn't necessary or useful. – Nathan Tuggy Jan 29 '17 at 21:52
  • @NathanTuggy But if mathjax is enabled, this way is obviously superior – suomynonA Mar 6 '17 at 4:57
  • 3
    The MathJax way doesn't allow you to do anything you can't already do (in particular, it hasn't the slightest effect on the ability to edit comments), is less clear than some other methods, and requires a dependency that will consume noticeable CPU power when rendered and is only available on certain sites, which increases mental load to remember two methods. I don't see any way in which it is anything but inferior. – Nathan Tuggy Mar 6 '17 at 5:41
  • It is not less clear, and has negligible impact on mental load. It is superior in the fact that it can be used in comments. – suomynonA Mar 6 '17 at 23:42
  • 2
    MathJax can be used in comments. The MathJax edit trick can't, because comments have no minimum-length edit requirement in the first place, so this answer's rationale for superiority appears to be missing the point of the question, which is to allow edits (to others' posts, with less than 2k rep) which would otherwise be prevented because they're too small. The given MathJax trick can do that, but only in a strict subset of the conditions that every other trick listed can: only on MathJax sites. It can't change the fact that the trick is only meaningful in posts. Not comments. – Nathan Tuggy Mar 6 '17 at 23:58
  • But it is a perfectly valid alternative method; also comments have a 15 character minimum requirement; this trick allows you to bypass that. And I know that isn't the main point of this post but that doesn't mean this trick doesn't work. – suomynonA Mar 7 '17 at 0:06
  • 2
    Both are true, but that's why I consider this method strictly inferior for this question, which isn't about (comment) posting minimums, only edit minimums. – Nathan Tuggy Mar 7 '17 at 0:08

You must log in to answer this question.

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