Here on Meta, we have a tag.

It contains many topics that describe in high-quality detail how the SE network functions. By convention, we mark these posts Community Wiki and protect them. This lowers the bar for community contributions, but raises the bar for would be spammers.

I propose that we do the same for "Answer Your Own Question" posts on the main sites, automatically.

I have been interested for awhile in finding ways to encourage Canonical Posts. What is a canonical post? It is basically an answer to a frequently-asked question, the FAQ form of a Stack Exchange problem/solution. Canonical Posts are especially useful at Stack Overflow, which is heavily weighted towards very specific solutions to very specific problems. The purpose of SE is not only to answer people's questions, but to collect information that is useful to others. Canonical Posts can provide a venue for more general solutions, applicable to a wider audience.

For a perfect example of a canonical post, see here:

How can I prevent SQL injection in PHP?

Making these posts CW from the start would quell the most common objection to these posts, which is rep whoring. While I personally feel that users who take the time and effort to write such posts should be rewarded for it, writing good canonical questions and answers is hard, and I suspect that the ones who succeed at writing good Canonical Posts are more interested in getting the information out there in a well-written, easily consumable form than they are about rep gain. Perhaps a new badge that rewards well-curated canonical posts?

Setting Answer Your Own Question posts to CW, protecting them, and providing a badge for well-curated questions could be the Canonical Question platform that we've been looking for.

  • 5
    Is "rep whoring" really a valid objection? I think this idea would work rather discouraging. – Bart Nov 4 '13 at 19:41
  • @Bart: I added a clarification. Rep is for folks who want to write quick answers, not necessarily for writing a mini blog entry. The barrier that needs to be lowered is a community-imposed one, a bias against self-answered questions. – user102937 Nov 4 '13 at 19:43
  • 2
    The thing is that the canonicals are curated by the community and often posts are kept up to date or keep being improved by the community instead of the initial poster @Bart. Also +1 for the effort to create more canonicals – PeeHaa Nov 4 '13 at 19:43
  • 21
    @RobertHarvey: Huh? The people who wrote those mini blog entries put their time and effort to write an excellent answer. Why should they not be rewarded with the reputation they deserve? Why would someone writing a puny 1 liner on a popular question get hundreds of points in an hour, but someone who actually put some effort into it should not? – Madara's Ghost Nov 4 '13 at 19:45
  • 1
    @Bart: Self-answered questions (when they're not canonical) don't seem to fare well. – user102937 Nov 4 '13 at 19:47
  • 3
    Related: meta.stackexchange.com/questions/137353/… – Shog9 Nov 4 '13 at 19:50
  • 6
    @RobertHarvey: I agree (self-answered questions tend to attract ill-informed "get a blog" comments), but I think that's a problem that should be fixed rather than be accepted. Besides, someone could still get around this by posting the answer moments later, so this seems mostly like an inconvenience. – David Robinson Nov 4 '13 at 19:53
  • 2
    I do think that introducing a badge for "Wrote a Community-Wiki question with score X or more" is a very promising idea, but I think it should be separated from the auto-CW feature. – David Robinson Nov 4 '13 at 20:00
  • 1
    @DavidRobinson rather, "wrote a self-answered question with score X or more on both the question and the answer" – John Dvorak Nov 4 '13 at 20:02
  • 1
    Community editing is not the problem, really. CW is more of an effect than a cause. If it's a good post, and it gets lots of views, it's going to get pushed into CW anyway through answers and edits. The real problem is how to codify the legitimacy of such questions. Yes, "How do I prevent SQL Injection" is a valid question. No, "what have you tried" is not a valid response. – user102937 Nov 4 '13 at 20:24
  • 1
    I don't often downvote feature requests but this makes me sad, the most altruistic group; the self answers, who share their knowledge should not have the only reward they get removed – Richard Tingle Nov 4 '13 at 20:26
  • 1
    I'm not sure you premise is true, I've made 3 self answered questions (6 posts) and 5 out of 6 are upvoted (the remainding one at 0). If anything they've got a better response than my average post – Richard Tingle Nov 4 '13 at 20:39
  • 2
    Strictly anecdotally, @Bart: Question I know the answer of (and linked), as well as Is it generally frowned upon to answer your own question immediately? and Direct answer of own question lead to immediate close, but there's definitely selection bias there (no-one's going to come to Meta to complain that their SAQ was too well received). – jscs Nov 4 '13 at 20:40
  • 2
    I agree there should be more canonical posts, but anecdotally most of those I've proposed over the years have been rejected and closed by the community. Stack Exchange staff making a clear statement about their desirability would go a long way towards encouraging them - making them auto-CW feels a bit like they are discouraged (even though I know the opposite is intended). – Pekka Nov 4 '13 at 20:41
  • 7
    Incentivising self answered questions by removing all rep incentive seems somewhat counter intuitive to me. – Martin Smith Nov 4 '13 at 20:45

Won'ts answer to "Do moderators earn a salary?" seems to be appropriate here.

Taking away the possibility of earning a reward from someone who did both halves of making a high-quality contribution to the site is, to put it mildy, not a very good idea. If I go out of my way to post a clear question that is applicable to others, when I already know the answer and didn't have to share it with anyone, then I sure as kittens want at least a little pat on the back for it.

If my contribution is immediately snatched from my hands and made public property, why bother? I'll just wait till someone else asks the question and answer it then so I can get some recognition.

If you're worried about backlash to the tune of "This isn't a real question if you had the answer already, repwhore" or "get a blog, dork" then maybe the answer is a temporary, but automatically-applied post notice on the SAQ to the effect of


This fine fellow had the solution at the time of posting the question! That is A-OK here, even encouraged! Please vote appropriately, as if you were reading any other Q&A pair.

I already do this in comment form on SAQs that I see; I'd be mighty pleased if it were a system feature.

What is a canonical post? It is basically an answer to a frequently-asked question, the FAQ form of a Stack Exchange problem/solution.

I understand this connection you've made between Meta FAQs and SO canonical questions, but I think the character of the material is different enough; SO answers are, when we get right down to it, partly about being smart, clever, and showing off your ability to solve problems to your peers. Removing that recognition removes a major incentive for participation. (Actually, the more I think about it, the more I think a poster should get a little appreciation for taking the trouble to make a Meta FAQ, too.)

I also understand your concern that SAQs might often be too specific ("localized"), but I don't think that CW is the solution, despite its originally intended purpose. A broadly-applicable canonical question is most likely to come from an invested member who recognizes the need for such an artifact (and on a particular subject) and sets out deliberately to create one. And again, I think that member should be rewarded for that effort.

CW status would certainly invite collaborative upkeep of an established canonical post, but I don't think it will encourage its creation.

In the end, I think making canonizination take hold as a social phenomenon here is going to require a purpose-built system feature of some kind.

  • In the absence of CW, how do you encourage posts that are more general in nature and applicable to a wider audience, and less "I have this very specific troubleshooting problem" and "what have you tried?" – user102937 Nov 4 '13 at 20:06
  • No, I don't expect that. I expect a generalized problem statement with a generalized solution. What CW was originally meant to be, not the "get out of jail" card it turned out to be. – user102937 Nov 4 '13 at 20:08
  • 2
    I don't see how CW solves that problem, Robert. Are you envisioning a localized problem statment slowly being edited over time into a more general form? I would be surprised to see that happen. I think the drive for a canonical QA has to come from an invested member who deliberately sets out to create one, starting from scratch. And I also think there should be a reward for that behavior. – jscs Nov 4 '13 at 20:09
  • Sorry about the re-ordering. Deleted by mistake. – jscs Nov 4 '13 at 20:09
  • 1
    Could there be a middle ground of "reward the original author until their contributions to the post are < 50%"? – jscs Nov 4 '13 at 20:14
  • 2
    @JoshCaswell this is pretty much what CW already does – John Dvorak Nov 4 '13 at 20:15
  • 1
    ... in the sense that questions are automatically made CW after a certain number of edits or answers posted. – user102937 Nov 4 '13 at 20:15
  • True, @Jan, but Robert is right that CW status does lower the bar (even if only psychologically) for participation. – jscs Nov 4 '13 at 20:19
  • 3
    If my proposed post notice goes through, I insist that "fine fellow" be bolded. – jscs Nov 4 '13 at 20:41
  • 4
    Every time this comes up, I feel the same - we need more of a purpose built system feature to encourage and legitimize these types of posts. I keep looking at our tag wiki system and thinking ... why don't you do this for us, somehow, tag wikis? Just have yet to think of an implementation where they could. – Tim Post Nov 5 '13 at 4:21
  • 1
    "Fine fellow" is a sort of sexist assumption, innit? – David Robinson Nov 5 '13 at 5:42
  • 2
    @DavidRobinson: It's possible someone could read it that way, I suppose, but I believe that either gender can be a fellow at a university, college, or similar organization, and it's neutral as an adjective: "My fellow programmers,..." – jscs Nov 5 '13 at 5:59
  • Possibly "fine individual"? Same or similar feel, no potential sexist implications. – Ben Barden Nov 7 '13 at 17:27

I think this would work if there would be a delay, so that some amount of rep has been earned on them before the CW kicked in, which might make it work better for the smaller niche questions that never get that many votes. I've got a few gold badges from views (so probably a lot of googlers) of my canonical-type self-answered questions (excel, vba), but have never gotten that many votes from them. Of course, maybe balancing that with special badges would work.

  • The idea is that CW would kick in on q/a pair post. Not sure what you mean by "before the CW kicked in" – John Dvorak Nov 4 '13 at 20:16
  • @Jan Dvorak, I mean delay the CW until after the rep floor has been passed. I'll edit for clarity. – Lance Roberts Nov 4 '13 at 20:18
  • 3
    I'm intrigued with the idea of a special badge (maybe two, silver and gold) to reward self-answered questions that preform well in a mix of views and votes, mostly views. Perhaps something in addition to the usual view based badges. The more that we do to reward and encourage these types of posts, the more users will see them as legitimate, with hopefully less 'get a blog!' brow beating as the outcome. – Tim Post Nov 5 '13 at 4:18

You must log in to answer this question.