2

This comes up a lot for questions that are closed as not being related to software development. (What triggered the question was some of the answers for the Are Marketing Questions On Topic question on meta.

It seems to me that there is a logical problem here - many things might be specific to programmers, but probably usually aren't very, but you don't know that without knowing the answer to the question. For example, questions about specific marketing tactics (in this case swag). But it's also an issue for most work environment questions (which I'm willing to stipulate mostly aren't good questions) and pretty common for career questions.

But presumably the person asking doesn't know if this is true, or they wouldn't ask. And people closing the question might not know of any reason programmers are different than the rest of the world for attribute X, but it's hard to prove a negative, since a single valid example invalidates it.

For example, what if someone happens to answer a post saying that their recently completed research has found that programmers have a uniquely strong response to swag, but only in blue packaging? Or that they have ranked all professions by correlation between ambient noise and productivity, and programmers have the highest negative correlation of all occupations, including brain surgeons and orchestra conductors? (Or, to the probable joy of several recent posters, that a gene has been found that greatly increases programming ability but makes the bearer social awkward and physically repulsive?)

In practice, most questions closed as not being specific to programmers probably really aren't specific to programmers, or are only slightly so, but then what should the policy be? Should they just be closed on the theory that anybody with reasonable evidence that it is specific to programmers will be able to get it re-opened?

  • 5
    "Should they just be closed on the theory that anybody with reasonable evidence that it is specific to programmers will be able to get it re-opened?" -- that's basically how things work right now. We've had questions re-opened when someone came along with an answer. – Adam Lear Nov 4 '11 at 19:08
4

There are a few different types of questions where the test applies, most of which are straightforward:

  • Questions that a reasonable person would recognize has nothing to do with software development. These are questions like, "What color tie should I wear to work?", "What do I do if my co-worker smells bad?", "What's your favorite member of The New Kids on the Block?". These are just categorically off-topic.

  • Questions that a reasonable person would consider off-topic, but are scoped to include an "as a programmer" clause. These are questions like, "What is your favorite pickle (as a programmer)?", "As a programmer, do you think Justin Timblerlake is dreamy?", or "What do programmers look for in a spouse?" They almost always try to ask about the mythical programmer "lifestyle", and are also off-topic.

  • Questions that are unambiguously about software development. Stuff like, "How should I optimize my Towers of Hanoi solver?", "Why do I have to use a sigil in PHP?", or "Is reflection the appropriate way to get access to a private method?" These are categorically on-topic.

Most questions will fall into one of the above categories, and there isn't a lot of thought required to determine if a question falls into one of them.

The one that gets the most pushback is the second group: it's a general question, but they want to know how a programmer would answer it. There's also the gray area you mention, where the asker doesn't know that the problem isn't specific to programmers.

Stack Exchange is a little bit different from other Q&A sites in that the asker's intent is secondary to the value of the question for the rest of the internet. We want questions that are going to help future visitors and attract experts to the site so they can help with the really tough questions that haven't been answered elsewhere. We don't want questions that are only going to help a single person or are going to make experts roll their eyes.

So it's important—for the sake of building up a corpus of reference questions—that what the asker intended or what the asker believes is taken out of the picture when evaluating a question's merit. Instead, does the question, as written, invite answers from a programmer's perspective? Is the problem really unique to software development?

That is, borderline questions like this are judged by the community's reaction to the question: in comments, flags, close-votes, and answers. Tells include:

  • None of the answers mention the programmer perspective
  • The question gets a ton of down-votes in a short period of time, or settles on a negative score
  • The comments on the question are argumentative and/or contain an number of clarifying questions for the poster
  • The question received a number of close-vote flags (this tell is restricted to moderators and 10k+ users)

If the community, through its activity, doesn't consider it to be a problem unique to programmers, it really doesn't belong here. It's kind of annoying not to be able to determine on borderline questions whether it's going to work ahead of time, but it's incredibly hard (if not impossible) to create a set of rules or guidelines that's going to cover every possible question or scenario.

Of course, as Anna mentions, if it can later be shown the problem is specific to software development, the question can be reopened: closure isn't an automatic death sentence. Usually this is prompted by a revision the question to demonstrate why it's a programmer problem (more than adding a "as a programmer" clause) or from creating consensus here on meta.

  • If the question contains the words "as a programmer" - that's another tell ;) – ChrisF Nov 5 '11 at 9:23
  • Actually, the greyest grey area that I'm talking about, is where a moderator (really anyone who can vote to close) wouldn't necessarily be able to tell if it's specific to programming. Also I was pointing out the logical problem of proving a negative (that it is not specific to programming), you never know if a study will come out showing that in an extensive randomized survey, no programmer had ever thought Justin Timberlake was dreamy. Your criteria are pretty good, and I suppose having it closed can be considered a form of answer: "programmers are not atypical for your question." – psr Nov 7 '11 at 17:26
  • @psr This isn't about proving a negative: if a reasonable person would think that the question isn't specific to programming, that's all that's needed. Only the obstinate and pedantic would seriously make a case that a question about Justin Timberlake being dreamy is the type of question that's on-topic here. There are going to be cases where reasonable people disagree (e.g. the swag question): in those cases, talking it out is going to produce consensus (in this case, the question's off-topic). – user8 Nov 7 '11 at 17:35
  • I just noticed that you didn't even include as a possible category of question: Questions that are on-topic and include the phrase "as a programmer". I do think there are some ways in which programming is different than other occupations, and a question about those differences could be on topic. It's a little strange that you don't recognize that even as a possibility. I guess you said "most questions". – psr Nov 7 '11 at 18:56
  • @psr If the question is about software development, it's on-topic: I fail to see how this isn't addressed. Perhaps rather than continuing to talk about this in the abstract, you can start listing questions that you have in mind? – user8 Nov 7 '11 at 19:31
  • @psr When a question attracts programmer-specific answers, it is fine. If a question includes the phrase "as a programmer" but only attracts general answers then simply stating "as a programmer" isn't enough to make that question on topic. – Adam Lear Nov 7 '11 at 20:43
  • @MarkTrapp - Here is a specific example programmers.stackexchange.com/questions/120141/…, in which such information as exists does show programmers with IQs a bit above normal, but I can't post the answer because it's closed. – psr Nov 16 '11 at 22:29
  • @psr While there's a case to be made this was a borderline question mentioned in my answer that was handled by community consensus (deep negative score, most close votes from regular users, negative comments), whether programmers have high IQs really has nothing to do with software development (falls into the programmer "lifestyle" set of questions). There is no problem in software development solved by knowing the answer to that. – user8 Nov 17 '11 at 6:01

You must log in to answer this question.

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