What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 128 Stack Exchange communities.

Searching on Google I found the question "Java : File.toURI().toURL() on Windows file" with an accepted answer that contains the unfortunate misunderstanding that RFC3986 forbids "file" URIs with an empty host component (e.g. that file:///path is an illegal URI):

Since file URIs have no authority segment, they're forbidden from starting with //.

I think users expect that the information in an accepted, up-voted answer on Stack Overflow is true. If accepted, up-voted answers, which will likely be ranked high on Google and considered authoritative by many, contain false information it will misinform people and risk undermining the credibility of Stack Overflow.

I thought that fixing this information bug would improve the quality of information found in the internet ever so slightly, perhaps preventing future bugs from being introduced due to a wrong understanding of URIs, so I tried to fix it, but the fix was rejected. Maybe because it changes the meaning of the accepted answer.

So I guess I went about it the wrong way. What is the correct process for fixing wrong information in an accepted answer?

share|improve this question
See also (not a dupe, but related reading): meta.stackexchange.com/questions/78438/… –  Matt Jan 8 '13 at 12:58
@Matt Thanks. I guess "Edit away and fix stuff" did not work in this case. I would rather not downvote or flag it because the answer is basically right, but just has a piece of false information in it. I think the answer is to "hope people read comments" and to make users aware that the big green tick may be displayed right next to blatantly false information. –  glerup Jan 8 '13 at 13:27
@glerup, flags aren't necessarily bad; they just mean "hey, somebody please look at this". If you flag as "other" you get a few hundred characters to explain why. –  Monica Cellio Jan 8 '13 at 15:45
add comment

2 Answers 2

up vote 10 down vote accepted

Proper course of action is to earn 50 reputation points on the site then leave a comment to the post author and let him fix his answer.

Radically changing answers is not acceptable, even if it's for a good cause. Accepted or not accepted is not really relevant, however you can leave a comment (after gaining 50 rep) to the question author telling him the answer he chosen is wrong and he might unaccept it.

In case of wrong answer, downvote and in case of extreme bad or misleading answer you can flag it as such and it might get deleted.

share|improve this answer
Isn't OK to flag an answer just because contains misleading information. You can flag it if it's not an honest attempts to answer to the question, but if it's wrong just downvote and leave a comment. –  Ramy Al Zuhouri Jan 8 '13 at 12:22
Biiiiiiiiiiig emphasis on the extremities required to warrant a flag-for-deletion; unless the post includes military-secret information by accident, the flag will get rejected for "flags should not be used to indicate technical inaccuracies, or an altogether wrong answer". –  Matt Jan 8 '13 at 12:22
@sha I disagree that it's irrelevant that an accepted, up-voted answer contains false information. A lot of people will come from Google and see a well-authored answer with a big fat green tick. And they will be misinformed and the misinformation will be there for years. I can leave a comment and ask the author to fix it, but if the author does not want to edit his answer, or if he's not active any more, or it can for some other reason not be fixed, then should the false information just continue to be disseminated forever? Wouldn't it be beneficial to have a way to correct false information? –  glerup Jan 8 '13 at 12:36
@glerup I mean not relevant for allowing radical change. Such thing is not acceptable in my opinion. If you see the original author is not responding to comments after a while and you're 100% sure your information is correct and accurate bring it here on Meta and 2K+ members can edit the post themselves. –  Shadow Wizard Jan 8 '13 at 12:39
@Ramy several of my own answers (not on the main site) got deleted just for being totally wrong and/or inaccurate. –  Shadow Wizard Jan 8 '13 at 12:43
In what site? That's not the way of doing it in SO. Usually the author of the answer deletes it if sees too many downvotes. –  Ramy Al Zuhouri Jan 8 '13 at 12:48
@Ramy it doesn't really matter, my point is that in extreme cases when the author is not deleting it's better to delete such answer. Pretty sure there's discussion here on Meta about it somewhere. –  Shadow Wizard Jan 8 '13 at 12:50
Spam, not welcome on our community, not an answer, are you aware that this doesn't fall in any of these 3 options? And afterall, if some people judge an answer right (maybe because has upvotes and it's accepted), and some other people judges it wrong, why deleting it? Who decides? Your way of thinking is just wrong. –  Ramy Al Zuhouri Jan 8 '13 at 12:54
In addition, don't forget that you can leave an answer of your own. –  Bill the Lizard Jan 8 '13 at 13:12
@Ramy I agree that this particular answer should not be deleted, since it is overall a well-authored answer that happened to contain a piece of false information. But that being said, I think that whether a particular URI syntax is forbidden is not really a matter of opinion and instead of spreading false information and Stack Overflow should have a process that enables its community to ensure that the information next to the big green check-marks, that a lot of people will be directed to from Google, is correct. –  glerup Jan 8 '13 at 13:12
@Sha False information is not acceptable in my opinion. And it will continue to be disseminated through Google and misinform and mislead the people who find it and trust it. They will introduce bugs in their programs and the world will be a worse place. I will not accept a change process where i have to comment, wait a while, ask somebody else (who may not know about the particular subject) to change it and hope that they will bother to fix it. I want to just be able to contribute a fix, have it peer reviewed for correctness (possibly also by OP) and be applied. –  glerup Jan 8 '13 at 13:20
@glerup but who says you are right and he is wrong? I don't want my posts to be changed dozens of times by anyone who think he knows better. If someone think I'm wrong I will gladly read his opinion and change my answer if needed. But letting everyone change everything is way too much in my opinion. –  Shadow Wizard Jan 8 '13 at 13:41
@Sha If the change is peer reviewed before it is applied I can't see the harm. I think it is much more likely that a bunch of people checking a fact will accept a correct change than one particular person, who wrote the false information in the first place. Also I think it is much more important that the information is correct than hurting your pride by editing your posts. –  glerup Jan 8 '13 at 13:59
add comment

First of all, the edit queue is for newish users. By intent, this gives you / them an opportunity to make simple, uncontroversial, changes to posts. This case doesn't fall into that category. You can't fix everything; if you see an inaccuracy, you should just leave it for someone with more rep.

For people with more reputation, there are more options. When in doubt, leave a comment for the author of the answer, and let them make their own repair. If you are absolutely sure that you are looking at minor mistake, you can fix it. The author can aways roll it back.

If the entire answer is, as far as you can see, completely and gloriously wrong, the best thing to do is to comment, downvote, and/or offer a better alternative answer.

Answers do get deleted by some of us, but generally there has to be a combination of general poor quality with wrongness. A coherent, well-expressed answer that is a recipe to tell the reader to format his or her hard drive (not the desired operation), well, that might very well elicit a deletion.

share|improve this answer
In this case deleting the answer is inappropriate. I simply wanted to fix a piece of false information. I understand now that the options are: (1) convince the author, (2) use a cumbersome process to have somebody else fix the mistake, or (3) accept the misinformation. What I take away from this is that on Stack Overflow authors' ownership of their posts are more important than ensuring that information found next to the giant green tick is actually true. I just have to accept that. –  glerup Jan 9 '13 at 9:39
add comment

You must log in to answer this question.

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