85

Context:
Gitolite V3 just got released (an authorization layer in Perl to manage read/write access on Git repositories).

As a result, all the old V2 documentation links I used this past two years in my answer are invalid (not only their content is no longer up-to-date, but the links are now 404).

That means I resolved myself to update 51 on my answers (as well as a few answers from other contributors on Gitolite).

... except I got stopped in my tracks with the following message:

You have already edited 5 of your own posts today (not including very recent posts).
Further edits are not allowed until tomorrow.

5 edit limit

I understand the need for that limitation, as explained in "Getting “Too many edits […] Further edits are not allowed until tomorrow”" by Marc Gravell.

But I would like to know then how that limitation is compatible with a large scale edit in order to update one's old answers. It would take me 10 days to complete my own updates... which seems a bit depressing.

Some suggestions (just for the sake of this discussion)

  • allow those edits when done on the same tag?
    (but not sure if it is compatible with "prevent gaming")
  • allow those edits if the reputation is above a certain level?
    (if a user is a "trusted one", > 20K, maybe this could be considered?)
    (maybe 40K on SO, due to the large number of users already at 20K?)
  • flag a moderator to ask for a reprieve?
    (for 24 hours, you are authorized to do as many edits you want)
  • ? Other suggestions?

Note that the edits I have to do are NOT simple copy-paste, replacing old links by new ones.
I have to read again the content of those old answers, assess their validity in the light of the new Gitolite release, rewrite some of said content, and then update the links.

This is NOT about trivial edits, this is about maintaining and improving a large set of old answers, and this Stack Overflow "limitation" is in the way.
Seriously, how many high rep users (>50K, >100K) have been exhibiting bad edit behavior?
Let dedicated users do their work.

Improve this question
22
  • 40
    +1 Trusted users should be trusted to edit their own posts without malice (see what I did there?)
    – Oded
    Apr 18, 2012 at 12:21
  • That is killing me. I have a few others edits to do (not even related to Gitolite) to other old answers of mine, and I now have to wait until tomorrow... this is rubbish.
    – VonC
    Apr 18, 2012 at 13:26
  • 2
    Whoa, didn't even know this existed :/
    – Manishearth
    Apr 18, 2012 at 13:53
  • 2
    Note that others have worse problems. I'm carrying out a tag cleanup on P.SE, with a self imposed rule that I should do maximum five posts at a time(to avoid flooding the main page). With 200 pages in the list (not all are copied to the CW answer), this could take awhile.
    – Manishearth
    Apr 18, 2012 at 13:58
  • 11
    The edit limit is a protection against ragequits, so it can't be waived for users no matter how large their reputation. (Has there been a >20k rep ragequit? I don't know but I wouldn't be surprised. Sure, there hasn't been a >500k rep ragequit… yet, but reputation is not an indication of who is going to ragequit. There has been a ♦ ragequit, come to think of it (no, I won't give details).) I know this sucks, and the limit could perhaps be raised (I like the idea of tying it with a tag, but it would be hard to implement, because you'd have to track retags to avoid a trivial workaround).
    – Gilles 'SO- stop being evil'
    Apr 18, 2012 at 22:46
  • 8
    @Gilles: the question though (about ragequitting) is: what is the problem? Why prevent high-rep users of improving the site because of a few bad behavior? Ie how many (with rep >20K) have raged-quit before? If we are talking about, say, less than 20 people, that "edit limit" (again, even for "trusted users") seems a solution that does more harm than good.
    – VonC
    Apr 19, 2012 at 4:10
  • 6
    As a complete aside you should probably complain to the Gitolite people that they've broken all their links and ask them to set up sensible redirects.
    – Rup
    Apr 27, 2012 at 8:24
  • 1
    @Rup: the gitolite case is just an example of a more general issue ("link rot"), hence my new question (meta.stackexchange.com/questions/130500/…), in the light of the new mechanism planned by SO (a comment made by community asking the author to edit when a broken link is detected)
    – VonC
    Apr 27, 2012 at 8:32
  • 7
    There were 88 posts pointing to http://download.oracle.com/docs/cd/E17477_01/* which is broken, I've edited 45 of them, but the remaining 43 posts are mine. And I can't edit them due to this insane limit. And I don't want to spend 9 days on editing them and in the meanwhile not being able to edit other old posts where necessary.
    – user138231
    Feb 28, 2013 at 14:25
  • 2
    @Chichiray Agreed. For me, the simplest solution would be raising the number of allowed edits with the rep. How many vandalism instances there has been for user above 100K? 200K? (300K in your case) What such a limitation solves? Instead, it prevents dedicated users to do their work.
    – VonC
    Feb 28, 2013 at 14:43
  • 5
    @Troyen how many? how often? Do you really need to block all high rep users who has real maintenance work to do because of a few bad apples? I agree that putting the rep limit too high won't change much for 99% of the users... but I am hoping that this page is the opportunity to reassess the problem (vandalism) and its validity, and to define a better limitation that doesn't penalize users of good will. By the way, that limit might be different on SO compared to SE sites.
    – VonC
    Mar 1, 2013 at 9:32
  • 1
    I'm not sure about the rest of the community, but I do for sure not post low quality oneliner or code-only answers without any further elaboration/explanation in detailed layman's terms. Practically every answer I own contains a link. A lot of those links are links to authoritative resources/demos supporting my answers. Whilst those links do not make my answer useless when they go down, they do break the overall quality of my answer. And I really, really care about the quality of my answers (and inherently also the quality of Stack Overflow which is only decreasing lately).
    – user138231
    Mar 6, 2013 at 14:45
  • 1
    @Chichiray and remember, as I have edited in my question, this isn't just about links. Sometime, the answer itself needs amends. I really hope that somebody from the dev team will have a look and at least provide some kind of official answer to this.
    – VonC
    Mar 6, 2013 at 14:49
  • 1
    Why would you be allowed only 5 edits on your own questions, but an unlimited on others?
    – Cole Tobin
    Mar 14, 2013 at 5:02
  • 1
    Killing your bounty, @Chichiray, since you already have one on the related feature-request. But yeah, this is done.
    – Shog9
    Mar 15, 2013 at 0:04

6 Answers 6

Reset to default
14
  • allow those edits if the reputation is above a certain level? (if a user is a "trusted one", > 20K, maybe this could be considered?) (maybe 40K on SO, due to the large number of users already at 20K?)

As Kevin mentions here, this limit should now scale with reputation. Folks who are posting many answers every day for long periods of time might well need to make more edits to old ones, so the system should recognize this.

Improve this answer
2
  • Sounds great, thank you Shog9, and thank you Kevin. Will we have more specific details on this limit? Will it be the same on Stack Overflow and for all the other SE sites?
    – VonC
    Mar 15, 2013 at 6:19
  • Same everywhere. And you should never ever hit it on SO.
    – Shog9
    Mar 15, 2013 at 15:08
46
+50

A solution I suggest would be to route edits > 5 into the suggested edit queue, and let whoever can see that these days vet them.

Obviously users who can access the suggested edit queue should not be able to vote on their own edits (just like you can't vote / see your own tag excerpt and tag wiki suggestions).

Improve this answer
9
  • Why not? Will I be able to vet my own edits? (since I have vetting privileges already, wink-wink...)
    – VonC
    Apr 18, 2012 at 13:05
  • 5
    Obviously users who can access the suggested edit queue should not be able to vote on their own edits (just like you can't vote/ see you own tag excert/ wiki suggestions).. I swear that was always there. I swear.
    – Matt
    Apr 18, 2012 at 13:09
  • 4
    Well they could automatically get one "accept" vote so they'd just need one more to be approved.
    – ThiefMaster
    Apr 18, 2012 at 16:41
  • @Thiefmaster: I believe that SE sites only require one approval to be accepted.
    – Matt
    Apr 18, 2012 at 22:22
  • @Matt: They required two as of this afternoon when I voted for a few edits (on SO).
    – Flimzy
    May 8, 2012 at 6:46
  • @Flimzy: This is definitely the case for SO, but I think the SE 2.0 sites (basically all the others!) only require 1 vote.
    – Matt
    May 8, 2012 at 8:54
  • 1
    This nicely address the account vandalism problem, but not the frontpage spamming problem.
    – Florin Ghita
    Feb 28, 2013 at 14:40
  • 1
    Not that I know the rules, but possible solutions to frontpage spamming include 1) if the question has an accepted answer and either are edited, they shouldn't spam the front page, or 2) if the total (+/-) votes applied to the question and all answers exceed a certain number, a spam shouldn't take place, or 3) add a checkbox for "This is a minor edit" on the edit page which, when checked, prevents a spam. None of the above qualifiers should prevent a review. I'm sure someone will tell me how irrational each of these ideas is.
    – JoshDM
    Feb 28, 2013 at 19:18
  • @FlorinGhita What if you were only allowed 5 edits + 5 suggested edits a day? Still twice as fast as now, but hopefully slow enough to avoid vandalism and too much spamming.
    – Troyen
    Feb 28, 2013 at 21:19
30
+1000

I would suggest the number of edits permited should be 5 + floor((number of posts)/100).

This will address the account vandalism problem as only one hundredth of posts may be edited. And also permits to very active users to edit more, proportional with their activity. (And is simple to code/understand it.)

Improve this answer
6
  • Why not? That would mean 75 edit for me per day, that should be enough :) +1
    – VonC
    Feb 28, 2013 at 14:57
  • I have 396 posts, so, for me would be 5 + 3 :)
    – Florin Ghita
    Feb 28, 2013 at 15:01
  • That's a great idea
    – Hugo Dozois
    Feb 28, 2013 at 15:16
  • 1
    1) Part of the idea here is to prevent ragequit vandalism; this doesn't address that. 2) Another issue this is to address is vandalizing an account that has been compromised; this doesn't address that.
    – Servy
    Feb 28, 2013 at 19:09
  • 4
    1. The user will be able to edit only 1 percent of his posts. This does not compromises the account. The damage will exists, yes, but will be limited. This is the ideea. 2. The same with vandalizing: Only one percent will be damaged. Now, sure, the 100 can be 200. The ideea can be implemented a little different or in combinations with other things.
    – Florin Ghita
    Mar 1, 2013 at 6:41
  • Perhaps a better solution to vandalizing or ragequitting would be to support a mass-revert from the profile page.
    – Bob Stein
    Mar 15, 2017 at 16:30
17

A workaround would be:

  • Log out (or, spawn an anonymous browser instance).
  • Make as an anonymous user suggested edit via "improve this answer" link.
  • Log in (or, switch to the logged-in browser instance).
  • Note the "An edit was suggested to your post" notifications in your global inbox. Review and approve them all. As you're the post owner, it's immediately approved without excuses. This is also insensitive to the 20 suggested edit reviews per day limit.
  • Profit.

Again, this is not a solution, but a workaround. This problem still needs to be solved in such way that one don't need to workaround it like that. This workaround is in essence an abuse of the suggested edit system.

Improve this answer
5
  • 1
    stackoverflow.com/review/suggested-edits/1704035 :)
    – gnat
    Mar 12, 2013 at 22:27
  • Intersting detail is that all other approves after the limit of 20 aren't logged in stackoverflow.com/users/157882/… I've done about 50 anonymous suggested edits the last half hour.
    – user138231
    Mar 12, 2013 at 22:34
  • 3
    I've just edited another >80 outdated links in my old posts within a hour. I love this workaround so much that I refuse to post a bug report that the 5 edits per day limit to prevent ragequitting is simply severely broken.
    – user138231
    Mar 13, 2013 at 3:07
  • 5
    This is the smartest workaround I've ever seen to anything ever.
    – Adam Rackis
    Mar 13, 2013 at 14:52
  • :)))! I like how easy is this workaround.
    – Florin Ghita
    Mar 15, 2013 at 6:56
8

Another solution could be increasing the number of allowed edits, basing on the reputation, which is what happens to the number of flags a user has in a day. The number of flags for moderation attention changes from 15, in a site where my reputation is 1317, to 45 on this very site.

For example, it could be increased by one every X points gained after 10000, or it could be defined basing on the number of answers with a score of at least 3, where 5 is the default value for users who are not 10k users.
Users with a higher number of answers have a higher chance of updating their answers; the fact the answer has a score of at least 3 means the community has seen the answer positively, and this would be a reason more for keeping the answer updated. Clearly, the number of answers with a score of at least 3 should be used to define the number of edits per day the users can do, not the answers that could be edited.

10k users probably need to keep update more of their answers, and 5 is probably a too low limit. I think there should be still a limit, as nobody needs to update a quart of the given answers in a day, considering that the limit is for not recent posts.
Increasing the limit doesn't mean the limit is necessarily reached all days. As far as I recall, I hit the limit in very few cases; I would have probably hit the limit once more, if it was higher, but that doesn't mean I would keep editing my old answers just for the fact I can edit 10 of my answers, instead of 5.

Improve this answer
-3

You can comment on all 51 answers saying "this answer needs to be updated in light of the new release (link) and I will be doing so by (ten days from now)." Then do each answer (as you say, it's a fair bit of work), 5 a day, and as you do an answer, delete the comment.

Situations like this where a single incident require dozens of answers to be changed are rare. Some might even argue that the old answers need nothing more than a comment that says "note this is for version x.y" and a new answer added that covers version x.y+1 or whatever. (If many people still use the old version, this is a good approach.) So I wouldn't support writing code to handle it. High rep users ragequit too. Handle these edge cases with the tools that are already in the system: comments, and the ability for one user to write two answers.

Improve this answer
11
  • 4
    This is not a solution. This is a workaround. And one with potentially a very large waiting list. I understand the concern of ragequitting, but being able to edit only 0,035% of the answers instead of e.g. 1% makes really no sense. I don't see how being able to edit 1% and having 24/7 active moderators with a ban-button would make ragequitting a severe problem.
    – user138231
    Mar 6, 2013 at 15:29
  • 2
    High rep users ragequit too? really? In so many number that I have to be limited in my edit work?
    – VonC
    Mar 6, 2013 at 16:03
  • 1
    @VonC really - an example
    – gnat
    Mar 6, 2013 at 16:06
  • @gnat ok, one example and I cannot do my work?
    – VonC
    Mar 6, 2013 at 16:22
  • @VonC I know about Internet time and all that, but really? Ten days? You're acting like it's a year. Code sucks at distinguishing between unusual external events and anger - and anger is much more common. There are tools that will let you immediately keep people from acting on incorrect information, and ensure no harm is done by the edit limit. Calling them a "workaround" doesn't make them that, and spreading your work over ten days isn't "cannot do my work".
    – Kate Gregory
    Mar 6, 2013 at 16:27
  • 1
    @VonC okay, find 4 more examples here
    – gnat
    Mar 6, 2013 at 16:28
  • @gnat I don't care, I just want to do my work, and this limitation is in the way. I mean: 200K+ rep, that should give me some right to edit my old answers...
    – VonC
    Mar 6, 2013 at 16:28
  • @gnat: I interpret "ragequitting" as deliberatly deleting/editing all old answers into garbage (note that there's no apparent limit on deleting old answers, while there is one on editing old answers, expect that one cannot delete accepted answers). The mentioned users do not seem to have done that. They just asked for deletion of their accounts.
    – user138231
    Mar 6, 2013 at 16:37
  • @VonC well you asked about ragequitting high-rep users, I gave examples (though the way Chichiray explained further, my examples appear rather weak). As for the limitation you mentioned, well I have no objections if it goes away, I would benefit from this too (got couple outdated links in my answers that would be nice to fix)
    – gnat
    Mar 6, 2013 at 17:07
  • 2
    @gnat: They did not ragequit. They just kindly asked the site admins to remove their accounts. I do for life not see how that would be a valid reason to limit the edits on my 14K old posts to exactly 5 per day. Even if they actually ragequit, moderators would easily have blocked and rolledback the actions.
    – user138231
    Mar 6, 2013 at 17:36
  • @Chichiray understood, thanks. Agree - my examples don't qualify
    – gnat
    Mar 6, 2013 at 17:49

You must log in to answer this question.

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