25

So, let me tell you a story about how a series of individually well-intended steps led me to apparently become the scourge of Stack Overflow.

I was working on a small project, and I ran into an odd problem: the program compiled fine on two compilers, but the third one gave a completely unhelpful and perplexing error message. I searched the web for a while, and could not find a single mention of that specific error message in the given context. The code was not particularly complicated or unusual, either... "I'll ask on Stack Overflow", I thought.

I typed out the question, and got a few more ideas as I was doing so. I checked a few leads and after another 20 minutes of research, finally had my answer (a missing compiler switch), and continued on with my project.

Fast-forward to a few hours later, I was clearing out my browser tabs, and saw that I had already typed out the question. I already knew the answer... though, considering how long it took me to find it, I thought that surely posting an answer myself would help anyone encountering the error message in the future.

I made sure of the following:

  • The error message was the question title. This should put the question as the top result for a Google search.
  • The question contained a complete, self-contained example.
  • The full output from the compiler was specified.

And looking at "What topics can I ask about here?":

  • Yep, this is a specific programming problem.
  • Yep, this is a practical, answerable problem that is unique to software development.
  • Yep, this includes the relevant details. (In retrospect I also could've included the compiler versions. I did not suspect that they could be relevant at the time. Someone asked for them in the comments, which I then provided.)
  • No, I made sure that this was not a typographical error.

So, I check "Answer your own question". After all, the linked Jeff's blog post says it's "explicitly encouraged", what could go wrong? I post it and go on with my day.

Half an hour later, I check back on it, and to my great surprise, I see:

  • 2 close votes;
  • 9 downvotes (between the question and answer); and
  • 0 upvotes.

I don't think I've ever received so many votes (up or down) in such a short period of time, ever. What happened?

The most-upvoted comment is:

What is the point of your self-answered question? Do you intend to document every single error that clang can throw?

So, you can probably see what happened there.

In the end, after I clarified my intentions in the comments, the votes mostly evened out, I walked off with a net increase in reputation (upvotes give more points than downvotes remove; not that reputation was the motivator anyway), the close votes timed out, and my question is now the #1 Google search result for that error message. Mission accomplished!

All this happened two weeks ago (I took some time to cool off as at that moment I was hopefully-understandably somewhat shocked that my well-intended attempt to improve the website was met with such animosity). Though, looking back at it now, I'm still curious why this happened.

  1. Have I made a mistake? Did I miss or misinterpret a site policy somewhere? Was the question too obscure / trivial for this website? The flow of votes mostly reversed course after I posted my intention in the comments... but was this necessary in the first place? Should I include a justification with all self-answered questions from now on?

  2. It seems to me that there is an apparent disconnect on the opinion on self-answered questions between the website administration and the community. Obviously, Stack Overflow the website benefits from good self-answered questions - it's free content, thus more pageviews and ad impressions. However, from the community's point of view, it's easy to see it as a lame grab for reputation. Self-answered questions seem to be held to a much higher standard... which is simultaneously understandable and not. Is there an actionable point here?

16
  • 2
    Not a 'scourge' at all - now there is a direct resource of information for that particular issue. Nicely done!
    – user6613600
    Commented Aug 18, 2016 at 10:48
  • 1
    Ha, well, it certainly felt like that during those few minutes. Thanks!
    – Vladimir Panteleev
    Commented Aug 18, 2016 at 10:51
  • 1
    I have done a question-answer as well, and it was reasonably positively received (and got another good answer)
    – user6613600
    Commented Aug 18, 2016 at 10:52
  • 1
    The c++ tag is a special place, at least that's the impression I have from seeing posts here on Meta. They seem to be less forgiving than any other "community" on here.
    – user247702
    Commented Aug 18, 2016 at 11:10
  • 11
    The community in general has an unconscious bias against self-answered questions. Just a couple of days ago I'd handled yet another moderator flag that read "This user answered his own question." like it was some heinous offense or something, and the last time I'd handled such a flag was just two weeks before. You just have to get used to it.
    – BoltClock
    Commented Aug 18, 2016 at 11:14
  • 1
    Its a shame how that was handled. We always tell people to write good questions, assemble a self contained example, make sure they research first ... etc etc but suddenly when it happens but is self answered we downvote it anyways. Why?
    – Magisch
    Commented Aug 18, 2016 at 11:22
  • 1
    see also How to ask and self-answer a correct, high quality Q&A pair without attracting downvotes?
    – gnat
    Commented Aug 18, 2016 at 12:04
  • 3
    Now that things have turned around in a positive way you might want to remove the meta comments under your question and answer there. This meta post serves as their replacement.
    – rene
    Commented Aug 18, 2016 at 12:08
  • 2
    @rene: Done. Screenshot of the deleted comments for reference/archival.
    – Vladimir Panteleev
    Commented Aug 18, 2016 at 13:16
  • Just a guess, but based on the order of events in your story, since you typed out your question before you did your research, maybe your question got downvotes because it "does not show any research effort"? Personally, I think that because you self-answered, it's obvious that you did the necessary research, but maybe others decided to judge the question independently of the answer.
    – Don't Panic
    Commented Aug 18, 2016 at 15:04
  • 6
    @Don't Panic: "others decided to judge the question independently of the answer" That's actually how it should be. But it's absurd that every question has to have some semblance of "I used all of these search terms/I Googled the exact error message and none of them turned up anything" or paragraphs upon paragraphs of data in order to pass for "demonstrated research effort" and be considered a reasonable question for the site. And what about questions where the self-answer is said research effort? Say "I found my own answer and I'd like to share it" upfront and it gets edited out as noise...
    – BoltClock
    Commented Aug 18, 2016 at 15:40
  • @BoltClock Is that always how it should be? I agree with you about the absurdity of including those things in questions. It seems like a catch 22, though. If all questions should be judged independently of their answers, and the research effort of ones like these is really only demonstrated by the self-answer, then it's reasonable for people to downvote them. It seems to me that in order to encourage self-answered questions, people should take the self-answer into account and refrain from voting based on lack of research effort apparent in the question.
    – Don't Panic
    Commented Aug 18, 2016 at 16:46
  • 6
    Posting a question in the C++ tag is like looking down the barrel of a gun when you have a hang fire.
    – user1228
    Commented Aug 18, 2016 at 17:10
  • 2
    Stop closing this as a duplicate of meta.stackoverflow.com/questions/251874/…, the answers there focus on that particular self-answered Q&A, which apparently wasn't of high enough quality.
    – user247702
    Commented Aug 19, 2016 at 13:30
  • @Stijn answers in the dupe are focused on (poor) quality that is unfortunately typical for most self-answer attempts. I think you'd have better chance to reopen if you tag this one with specific-question to better indicate that discussion here is focused on a rare case of a particular question of reasonable quality
    – gnat
    Commented Aug 19, 2016 at 15:00

2 Answers 2

Reset to default
31

This seems like a perfectly valid question to me. In fact, I have upvoted it and the answer.

There is only one main caveat for self-answered questions like this (or, really, any question), and that is that you should only be posting things that contribute positively to the available body of knowledge. That means any number of things, but one of the most important things that it means is that the answer to it should not already be trivially found on the Internet or the official documentation.

It seems that, in this case, you ensured that was not the case before posting the question, which is good. And, as you put it:

my question is now the #1 Google search result for that error message. Mission accomplished!

Indeed. I have no idea why you received such a poor initial reputation. Chalk it up to the question being tagged , and therefore appearing on the radar of many experts who already knew the answer and deemed the question to be "too simple".

There is nothing wrong with basic questions, though, as long as they are well-asked, answerable, and do not already have a good answer that is easy to find. Yours didn't. Now it does. Everyone wins. Ignore the naysayers; you did everything right.

2
  • 1
    I agree with this wholeheartedly and have also upvoted both
    – user6613600
    Commented Aug 18, 2016 at 11:19
  • Thanks (and also for your answer edit)! I feel saner now :)
    – Vladimir Panteleev
    Commented Aug 18, 2016 at 11:28
-4

There is a different way of looking at this by following this order of events. Content has been deleted since this happened and I can't see that content so perhaps I'm just putting the pieces together badly here. If so - my apologies up front.

Here goes.

  1. You post an honest, good question + self answer. Good for you, good for all of us.
  2. Given the nature of a particular set of tags, people don't really take well to the content and initially downvote it and even close vote it. That's a freedom people have; nothing wrong with it. It stinks if your content gets downvoted especially if its just good content, but as others have already indicated SO too has its tougher neighbourhoods and this can happen. We don't have to like that, but what can you do about it when its all anonymous.
    • The close votes are really weird though, I can't argue with that. But close votes being cast for no good reason is also not something new.
  3. Someone asks you to explain yourself in a comment; that could have been worded a whole lot better, but to me the question seems a sincere attempt at getting more context. Bless them for giving you a platform to explain yourself to the critics, no matter how it is phrased. Perhaps the intent was no good after all but you can't really prove that so its just best to either assume the best or ignore the worst - and just respond.
  4. You did, thankfully
  5. Your added context makes the question seem a whole lot better and now upvotes start to trickle in, tipping the scale
  6. There is no longer a disturbance in the force

That is still nowhere near pretty as you shouldn't have to explain/defend the reason for your content to exist. But it worked out in the end because you did take up the challenge of the comment to explain yourself and that is still a win in my book.

Besides that... Perhaps I have rose-tinted glasses, but what I certainly don't have is a radar on which I can detect animosity behind downvotes. I'm just saying.

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