12

Recently, a guideline to demonstrate a "minimal understanding of the problem being solved" was made mandatory on Stack Overflow:

"Questions must demonstrate a minimal understanding of the problem being solved. Tell us what you've tried to do, why it didn't work, and how it should work. See also: Stack Overflow question checklist."

Despite this rule, I've found many popular questions that have remained open, even though they don't demonstrate an attempt to solve a problem before asking for its solution:

How to modify existing, unpushed commits?

What is the difference between 'git pull' and 'git fetch'?

How can I get query string values in JavaScript?

What is a plain English explanation of "Big O" notation?

Can comments be used in JSON?

How to resolve merge conflicts in Git?

Trim string in JavaScript?

Algorithm to find the most common substrings in a string

When posting questions, I sometimes follow the example that has been set by these favorably-reviewed (but not well-researched) questions, based on the assumption that it is acceptable to ask questions like these under some circumstances. Is it now forbidden to ask questions similar to these, despite the almost entirely favorable reviews that these questions have earned?

5
  • 5
    Aside from your own question, the other 4 are 4+ years old and asked when the rules of the site were different. Does it mean they should or shouldn't be closed? I'll let the community decide, but it is hard to evaluate "consistency" when you are comparing today's standards to older guidelines. Jul 9, 2013 at 16:08
  • 11
    Age of the questions notwithstanding, Stack Overflow is a big site within which only a sub-section will be fully aware of the site's goals, and only a small part of that is willing to enforce those goals even when the content is fun/interesting/popular, but does not fit. Popularity, as a result, is by no means a good indicator for the acceptability of certain questions.
    – Bart
    Jul 9, 2013 at 16:08
  • 4
    I don't think "Algorithm to find the most common substrings" is a bad question by this standard. (I hope you don't either, since you asked it!) Jul 9, 2013 at 16:10
  • 5
    For my part I always take the required research effort to be proportional to the required effort on the part of the answerer. If its an incredably simple question to answer I'd only require that the question be very clear. A more complex question requires more research. This doesn't seem unreasonable otherwise you're effectively banning basic questions which I don't think anyone wants as they will probably be useful to thousands of people Jul 9, 2013 at 16:11
  • I'm sorry that I posted this question here; I shouldn't have been so paranoid about the enforcement of these rules. Apr 24 at 20:19

4 Answers 4

20

I'm not entirely happy with that close reason for this very reason.

See, the intent here was to handle the sorts of "here's my spec, please write code for me" questions that were already being closed - not expand closure to damn thousands of existing questions with good, useful answers. For now, I've retired that OT reason and replaced it with:

Questions asking for code must demonstrate a minimal understanding of the problem being solved. Tell us what you've tried to do, why it didn't work, and how it should work. See also: Stack Overflow question checklist

To be clear: it's always better when a question implies that the asker knows enough about the subject matter to understand a reasonable answer. However, there are a fair number of questions where this is implied merely by the fact that the author knew enough to ask them. As Richard Tingle suggests,

For my part I always take the required research effort to be proportional to the required effort on the part of the answerer. If it's an incredibly simple question to answer I'd only require that the question be very clear. A more complex question requires more research. This doesn't seem unreasonable otherwise you're effectively banning basic questions which I don't think anyone wants as they will probably be useful to thousands of people

That's a good strategy, and one that fits well with Wikipedia's Assume good faith precept: if nothing leads you to believe the asker doesn't know what he's asking, work on the assumption that he does.

We'll continue to monitor the use of these reasons and tweak them as-needed.

4
  • 1
    Given this change, how do you deal with the idle curiosity/rhetorical questions?
    – user102937
    Jul 9, 2013 at 17:43
  • Rhetorical questions should generally be lumped into either "unclear" or "too opinion based", though I'm not sure that covers all the bases right now. The validity of curiosity questions is debatable - I think we'll want to discuss these on a case-by-case basis.
    – Shog9
    Jul 9, 2013 at 17:57
  • @Shog9 Recently, another moderator told me that this specific closing reason is intended for all questions that ask how to do anything, and not only questions that ask for code. I'm a bit worried about this trend, since it makes it easier for the Stack Overflow community to ostracize beginners. Dec 6, 2013 at 16:42
  • 1
    @Anderson: chime in here, please.
    – Shog9
    Dec 9, 2013 at 20:54
5

All of those questions you cite are 4 years old.

They were acceptable then.

Now, if they were asked, they'd likely be closed as a duplicate, or as requiring minimum effort.

Those questions are still open simply because they haven't been closed by the community. It only takes 5 votes.

As it stands, we (moderators) are not in the business of cleaning up old posts until we've got a handle of the tens of thousands of questions asked per day currently. I would not recommend flagging these with a custom flag ("Other"), rather, if you don't have the votes to close, you can flag it for closure, and then it goes into the Review queue.

If the community wants to go back and clean up old posts, that's up to the initiative for that group of people. But really, if it's not actively hurting anything, why does it matter?

Even your own question you cite(!) is still written well enough to 'get around' that requirement. It's got a good answer, it isn't a glaring eyesore, so it's still open.

If the community wants to close it, they can always vote to close it. If you want it closed, you can always flag it for closure, and it'll be put into the review queue. If the community agrees with you, it'll be closed.

3
  • 1
    Well, except for the "most common strings algorithm" one. I think the point should be made that we're not in the business of going back and cleaning up very old questions to meet today's standards.
    – user102937
    Jul 9, 2013 at 16:18
  • 1
    @RobertHarvey Nonetheless, I'm worried that the presence of popular (yet non-compliant) questions may be misleading to some users. For more than a year, I have followed the misleading example that has been set by these questions, and I don't want other users to repeat the mistakes that I have made. Jul 9, 2013 at 16:25
  • 4
    @AndersonGreen: Broken windows have been a problem for SO for as long as I can remember. :)
    – user102937
    Jul 9, 2013 at 16:31
2

In addition to what Cody said, I'd like to focus specifically on the last question you linked.

It's an algorithm question, and we've already had a discussion this morning about those. In general, these kinds of questions tend to be "softer" questions, but as long as they stay productive, I really have no problem with them.

The "questions must demonstrate minimal knowledge" close reason is really for icanhazcode questions, where the OP is asking for someone to write a complete solution for them. It compels the OP to demonstrate effort.

0

It is unacceptable to ask these sorts of questions, but the questions that you linked are very old. Back in the land of the dinosaurs, where you could do whatever you wanted and gain rep. The rules weren't as strict, and many people had these questions, so they upvoted and let it stay open.

With questions like those, I usually flag for a moderator. I leave a message similar to this

Should be locked for historical value, but it isn't a good example of what to ask. It shows little to no research value

Or you can flag for a moderator to close, since you don't have close votes yet

4
  • 3
    We only use historical locks under very narrow, specific circumstances.
    – user102937
    Jul 9, 2013 at 16:10
  • What might be these circumstances Jul 9, 2013 at 16:11
  • 2
    meta.stackexchange.com/questions/126587, under the subheading "When is it appropriate to lock a question for historical reasons?"
    – user102937
    Jul 9, 2013 at 16:12
  • 2
    The questions that are listed would be bad to flag for moderator attention (using a custom flag), for the reason Robert brings up: Historical locks are used very rarely, and we've even discussed not using them at all. As it stands, these questions should go through the community -- unless there's a real fight breaking out over the question, there's no need to involve moderators. Jul 9, 2013 at 16:46

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