This has beendiscussed, although I am not sure there is a consensus. But since the introduction of the new close reason, it seems like the closing system is a bit unbalanced, because a generic reason was replaced with a more specific, less widely applicable one.
Scope
This question is about "How to do X, I have no idea how to start" questions, such as requirement dumps, most often with no code, or the "I have this code, how do I make it work on even numbers instead of odd" sort, which are not "too broad" (as in "How do I write a Facebook clone"). I believe this is what is commonly referred to as "send me teh codez". I'd call them zero-effort questions.
The "official" solution
The suggested way to treat low-effort questions is to downvote / ignore them: if someone wants to waste their time answering this, the question may help someone else in the future. An important argument is that closing requires more effort from the community, and we instead want those questions to be unanswered, downvoted, and as such automatically deleted later, to reduce their cost for the community.
Why it doesn't work
The fact of the matter is that such questions are often closed anyway. From the newest closed questions search, here are some recent examples, with close reasons:
And from the Close Votes review queue:
Even though we are told that a close vote is not a super-downvote, apparently the higher-rep part of the community (those with close privilege) do not want such questions answered on SO and want to make it clear to the askers, too. This is, in a way, a response to Shog9: a lot of people are ready to put in the necessary effort, they are doing it anyway, but they are forced to choose the close reason rather arbitrarily, which is a bad thing.
It's been also pointed out that if a question has two answers, however bad, it won't be automatically deleted.
My stance
We should differentiate between low-effort and zero-effort questions. The former are just bad programming questions, let them be. The latter are not programming questions at all, and as such, are off-topic. That means that the asker is not a programmer, at least in the context of the question at hand. To have a programming problem, one has to do some programming first, in the most general sense of the word. This includes research, trying something, etc.
In the past I used the "lacks sufficient information" reason, but it's gone now. The reason replacing it is more specific, and I don't think it applies here (although I like it). Applying any other reason is also often a stretch. I want a better close reason (and the "permission" from the community to close).
The cost can be reduced by increasing the weight of off-topic close votes for gold badge holders, as it has been done for duplicates.
Also, if someone doesn't want to spend time closing these questions, then please don't, and the Review queue can now filter by type, so no one is forced to look at zero-effort questions there.
Critique
This is exactly what I suggest. I'm not saying the person is not a programmer or not a welcomed user for asking this. But in the context of the question, they are not a programmer, but a help vampire. And it's the question that is not welcomed, not the person.
- How can you tell if the OP has done research or not? Why call them names? You're not God!
I judge the post, not the person. If the question has no evidence of research, it appears off-topic to me.
- Any answered question can help someone in the future. Just leave those people alone!
Yes. It is possible that someone who give themselves an exercise of searching Google may find that somebody has done someone else's homework on SO, and actually learn from it.
But, while we're discussing the future, there are disturbing trends calling for measures, with evidence that the problem is related to declining question quality. Is ignoring extremely bad questions and hoping they go away really the way to go?
Also, I doubt that such questions are attracting many views anyway (although I have no data).
Alternative solution
An alternative solution has been suggested in the Podast #59, and boils down to: "Let's just display crap less often on the front page". I have no problem with that, and it may help save the overall impression SO makes, although it is not specific to a certain kind of bad questions. I say, do your thing, at let us do our thing.
TL;DR
Why close: because the newcomers don't even realize something is bad about their question. We need to let them know what the problem is. We don't want to reward them by answering. Also, we kinda dislike rep-whoring answers. Zero-effort questions are the most frustrating kind of bad questions on the site, their numbers are growing (or it feels like it), and we need to target them specifically.
Bottomline: people are closing them anyway. We should either do something to stop it (why, though?), or give them a more suitable close reason, like we did with "debugging help", or adapt the "debugging help" to encompass this case.
