I asked a very specific coding question on Stack Overflow. I am at a loss as to how to proceed, so I don't have a code sample. I immediately got a -1 and a complaining comment for not having a code sample. Is a code sample required to ask a question? The question is https://stackoverflow.com/questions/21004449/how-to-make-a-page-that-builds-and-submits-a-search-to-google-that-is-optimized

Would a different wording make it more acceptable? If so, what should I say?

  • 12
    SO is not a place for you to just post a list of requirements and get a completed solution. While questions don't always need to have code, they usually will. Note the comment is not saying that the question lacks code, but rather that it lacks any clear effort/research/attempt on your part. Effort/research/attempts need not necessarily mean "code", but that is generally the easiest/most effective. Also, just because someone writes a comment explaining a problem with your post doesn't mean they downvoted you; it's an assumption you shouldn't be making. – Servy Jan 8 '14 at 19:33
  • To add to whar Servy has said, usually if you expect code in the answers you should have your attempt at code in the question – Richard Tingle Jan 8 '14 at 19:35
  • 1
    So reading the Google documentation doesn't count as a research attempt? I guess I'll delete the question since no one is being helpful. – George Jan 8 '14 at 19:36
  • 3
    @user1981459 No it doesn't count as a research attempt. Instead, take what you "researched" and put that in your question. For example, "Google documentation says 1 + 1 = 3, but I don't know how to add." – Kermit Jan 8 '14 at 19:38
  • 3
    The question in question got a couple more downvotes almost immediately after this Meta question was posted. I think it's obvious user1981459 has already realized their question has problems (and is now asking for our help), piling on more downvotes won't help much. – yannis Jan 8 '14 at 19:39
  • Thank you, Yannis. At least there's one friendly voice here. – George Jan 8 '14 at 19:40
  • @user1981459 We are only trying to answer the questions you have posed – Richard Tingle Jan 8 '14 at 19:42
  • 3
    It's hard to be friendly when a small group of users have to go through 5,000 questions like this one. Help us by improving your question. – Kermit Jan 8 '14 at 19:42
  • Well, to be honest, if I had discovered your question naturally (through browsing the site) and not through this Meta question I would have probably downvoted it too. It's not really a matter of friendliness, downvotes are just a signal that there's something wrong with the question. That said, you've already got that, I don't think the additional downvotes serve any purpose. – yannis Jan 8 '14 at 19:43
  • 2
    @Yannis Really, because I read this meta question as saying he doesn't think his question has problems and that he feels that the feedback is wrong. Getting several more downvotes indicates that the community feels his post isn't of high quality, rather than that this one user is being mean. Also, a post's score isn't just for feedback to the author. it's for feedback to others looking at questions on the questions list, for the q-ban algorithm, for search results, the "interesting" homepage tab, etc. – Servy Jan 8 '14 at 19:43
  • 2
    @Servy I guess I'm at that point where if a Meta question doesn't invoke Godwin's law, I read it as an honest request for help... ;) Now, how about you post your comments as an answer, so I can properly upvote them? – yannis Jan 8 '14 at 19:48
  • Sorry, I'm not trying to say my question doesn't have problems. I'm having trouble understanding why my research attempt was insufficient. I read through the documentation and didn't find anything relevant. Does that mean I shouldn't ask the question or that it was poorly worded? Or are people assuming that if I couldn't find it I didn't look hard enough? – George Jan 8 '14 at 19:52
  • It's not enough to do the research. You need to show the results of that research in your question. – ale Jan 8 '14 at 20:35
  • @user1981459 You're right that the site provides absolutely no guidance on how to ask your question. – Kermit Jan 8 '14 at 20:49

A code sample is not strictly required, no. What is required is for you to explain where you're at, what particular bit you're stuck on, and (if possible) what you think you need to take a step forward. Since we're dealing with programming problems, it's often the case that a code sample is the quickest, clearest way to pass on a large part of that information. Trying something -- anything -- is also usually a good way to place yourself in the problem space such that you can explain your location to other people. You can absolutely do it in prose, though.

It looks to me like you're just way too early in your research on this problem to be posting a Stack Overflow question. You need a lot more material about your grasp of the situation. As it is, your question reads like a spec that you're expecting someone else to fulfill, which is not something that we do here.

See also How does "proof of effort" make a question better? and https://meta.stackexchange.com/questions/182266/how-much-research-effort-is-expected-of-stack-overflow-users/182380#182380=

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