What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 127 Stack Exchange communities.

Hi all, if this is a duplicate, I sincerely apologize. I wasn't able to find it when I searched, but that doesn't mean much.

There have been several occasions (this one for example) on SO where I have had a rough idea of the answer to the question but I haven't had enough information to give a decent answer. In the example, for instance, I knew what the OP was doing wrong, but I didn't remember the name of the method that needed to be called. I was able to give a high-level answer and explain what was going on, but I couldn't actually answer the question. As such, I posted in a comment rather than an "answer" because I felt as though my response wasn't actually answering the question.

So my question is, what's the accepted practice for that? Did I do the right thing or should I have posted an answer knowing it was incomplete?

share|improve this question
add comment

3 Answers

up vote 17 down vote accepted

Most of the time, if I don't have a concrete, specific answer, then I simply don't answer or comment - somebody else will take care of that. But if the question itself is vague, or it doesn't seem that anybody else is willing to answer, then I generally use the following three-pronged approach:

  1. State your assumptions.

    "This question appears to be about spronking a glibulet. You don't specify your framework, but based on the example code it appears to be Pippleblarg 4.0."

  2. Answer the question to whatever extent you're able.

    "If that's the case, you just have to make sure you set the MaxWidgets property to 42 before running the Churn method."

  3. Make explicit any known gaps/deficiencies in your answer.

    "This code is not thread-safe and may cause unexpected hyperfrobulation when used with PB versions prior to 4.0 or when Jupiter's orbit is aligned with Saturn's. Also, I haven't included any error-checking, which I would expect to see in production code."

It's pretty unlikely that people will downvote you if you explicitly define the "scope" of your answer, so to speak, and in many cases the person asking the question will be smart enough to fill in any gaps.

Oh, and one other thing:

  • If people leave comments, edit your answer later to fill in any missing information or correct any mistakes. Or, if somebody else posts a significantly clearer answer than yours, consider deleting yours to reduce the clutter.
share|improve this answer
I wouldn't delete your answers, but definitely upvote other helpful answers. One day the SO team will implement that sportsmanship badge, and when they do I will have been ready. –  Super Long Names are Hilarious Apr 4 '10 at 21:51
@Len: I did say consider. But definitely upvote other answers too. I would hope that the Sportsmanship badge will take questions into account where you deleted your answer, as that's really the ultimate in sportsmanship. –  Aarobot Apr 4 '10 at 22:40
I generally never answer I question that I can't back up with some reference (even just a little). This is off course sometimes annoying when you 'think' you know the answer, but can't find any resource that supports it. However, the quality and trustworthiness of the answer should be worth more than my own ego :-) –  Ivo Flipse Apr 5 '10 at 8:33
On some esoteric subjects where questions get very few views, I'd rather leave a partial answer as a comment after 24-48 hours than leave nothing at all. Better to give the person some pointer in a better direction than leaving them completely hanging. –  Michael Maddox Apr 5 '10 at 12:34
I've gone the comment route too, when my answer would be really vague and possibly incorrect but still better than nothing. It doesn't happen that often, though; most of the time when I need to leave disclaimers, it's because I found the question unclear or had previously answered/worked on a problem that was 90% similar but not exactly the same. Beyond that, there's a very fine line between know just enough to write a comment and don't know enough to respond at all - often better to assume the latter. –  Aarobot Apr 5 '10 at 15:28
+1 For awesome metasyntatic variable names. And a good answer too :P –  Cam Jackson Jun 30 '11 at 2:10
I am also a fan of Pippleblarg 4.0. It's some much faster than previous versions! –  Zoot Jan 6 '12 at 16:52
I'll spronk your glibulet, if you know what I mean :-P –  Rocket Hazmat Jan 6 '12 at 17:29
add comment

I guess I'm a lot less worried about getting downvoted than I am about someone not getting any discussion. I'll post an answer I'm less than certain about, and just state that. I'm ok with being wrong these days. If I have the time/ability to look it up to become more certain, then I will. Otherwise I can just edit/delete the answer later.

Hell, if I'm so drastically wrong that it motivates someone to show me up with the right answer, the asker STILL wins. ;)

share|improve this answer
add comment

I tend to do the same thing - add a comment at most - or not reply at all when I'm not sure.

On the other hand, it could be that what little you know already helps the asker enough to get the problem sorted. In exotic cases that are not likely to get answered anyway in the next few minutes, it may well be worth answering.

The only important thing then is to make totally clear what you're not sure, or unclear about. As long as that's transparent, I see nothing wrong with creating an answer.

share|improve this answer
add comment

You must log in to answer this question.

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