17

Nearly every day, I see a question I suspect of being a duplicate. Sometimes I find a dupe, sometimes not. Common problems are so familiar to many of us that we know whether a given question has been asked before. Nonetheless, it's getting harder to find them. It's clearly easier to answer some of them, the ones with short answers, which others do, some of whom are long-time users. Sometimes the new answer gets more upvotes than the answer to the original question it is virtually identical to. But that's how SE works, and I've been used to it now for years; so it seems more amusing to me than unfair.

I have gold badges in a number of popular tags, so often I can't cast a close vote without closing it. So I leave a comment, thinking either others will support by voting or the OP or someone else will object. It happens sometimes that a user will object on the basis of trivial differences in wording or context, even though the answers to the duplicate solve the problem.

Sometimes I just throw my hands up and hope the question fades away and doesn't come back to my attention. I feel there is not the same support for closing duplicates as there used to be. Maybe that's the way the community wants it. Maybe folks do not want to be involved in such moderation tasks. The tasks take time and they're not very interesting. Some folks seem to think closing as a duplicate gives the OP negative feedback, which they would rather avoid doing. If I remember, I go back to questions I commented on and vote to close, if appropriate.

I'm wondering if it's better to give the OP a specific solution than close as a duplicate and make them figure out how to change the variable x to y, or whatever adjustment might be needed. It's more interesting to write an answer, even if it's the same answer I wrote last year, than to search for an old dupe. And sometimes it's less work.

For context, the following prompted this outburst:

How do I pass a "list of lists" as the argument to a function of the form F[x,y]?

A couple questions away there's the same problem:

How to substitute values from a list into a function?

I suggested they are duplicates of this:

Using sets of arguments from a list

I don't mean to ask about these specific questions. My personal motivation is from wondering whether the effort and closures are appreciated enough that it's worth my time to do it. If the community does not feel that closing as duplicate is valuable to the site -- it's a valid reason, but I'm asking, is it a valued reason? Perhaps some are thinking we should mark duplicate less often. It gives the OP specific help and allows others to earn rep points.

So is closing as a duplicate still a valued close reason?

How exactly like a question should a duplicate be? (I think this is hard to define, and the first question is more important to me. However, I feel this might be part of the issue in some cases.)

| |
  • 1
    If you can straight-up search what the question is asking, then, perhaps yes, it should be closed, and folks should be redirected there. However, if the problem is similar but not exactly like what the OP has asked, then the duplicate should still be marked at the top, or edited into the question, but answered anyway. New methods are always popping up, and sure, they could possibly be relevant for the older question, but this is not a certainty, and version changes may negate the solutions some assume to "just work" for what the OP is asking on a newer version. – CA Trevillian Aug 16 '19 at 18:29
  • Related: 1841 with a relevant comment – Kuba Aug 17 '19 at 8:20
  • More reading mathematica.stackexchange.com/help/duplicates with a linked post that is very related to your questions: stackoverflow.blog/2010/11/16/… – Kuba Aug 17 '19 at 8:27
  • 1
    @Kuba I guess I'm just dense: I can't figure out which side Atwood is arguing with "use your good judgment." Does "duplication" mean (to Atwood) questions that are (or should be) closed as dupes or questions that are essentially the same that might or might not ought to be closed? In any case, it's an old article and I'm wondering if things have changed in our MSE community. – Michael E2 Aug 19 '19 at 14:52
  • I am very interested in sharing my view on that subject but I am short with time lately. Be patient :) – Kuba Aug 19 '19 at 15:02
  • I think it should be acceptable to post an answer by applying a known duplicate solution when one thinks that there is some added value, for example when: Is not evident for the OP how other answers apply. Other answers are buried in complicated context and more distilled version can be provided. The answer can consolidate many previous answers into one. Finding the other answer is not trivial and we need to make Q&A easy to search and find. One suspects the duplicate tag will be considered as unwillingness to help. Consequently should be OK to answer and voting to close at the same time. – rhermans Aug 20 '19 at 17:00
7

Slightly subjectively:

Surely some of us value marking duplicates more than others do. Otoh I'd be surprised to see that someone, who understood that such mark is not an equivalent to "use-a-search-engine!!!" comment, would end up directly against that model.

The duplicate header states "This question already has an answer here:" and this is how I approach resolution of that cases. The question will almost never be a duplicate, but it may have the same set of answers.

If conditions given for the question imply answers applicable only for the one of them then a better duplicate is needed or we can just link a closely related comment. (e.g. Times @@@ array vs. foo @@@ array) surely Times case will have many highly optimized approaches due to its nature. Don't forget about ability to add many duplicates to the same question to complement each other.

Remember that the answer Cases[list, _?Positive] can have maaany variations, you will probably fail to reproduce all alternatives people came up with in 2012, together with timings etc it makes quite a task.

Summing up:

  • Feel free to comment or even answer but I'd encourage to search for duplicates anyway.

  • You are not obliged to do so of course.

  • A measure of similarity between questions should be a similarity of sets of answers that are applicable to them.

Pros of closing:

  • most relevant answers / answers overview in one place or at least directly connected
  • the same question with different but good answers needs us to deal with merging that could've been avoided.
  • it is really hard to find a duplicate when you don't know the answer so help newcommers by making a connected net of differently phrased questions

Cons:

  • at first sight it could feel as an attack on OP from their perspective
  • it takes time

p.s. let's help each other with finding duplicates: Finding duplicates

| |
0

I agree with closing questions as duplicates, but to me duplicate should be interpreted strictly. Is the question slightly different? Then don't close it. Obviously this is subjective.

I do disagree with marking questions as duplicates and pointing to huge catch-all questions like What are the most common pitfalls awaiting new users?. That question and its answers is a great resource that people should read, but is far too broad to be a duplicate of any more specific question.

| |
  • Questions linked to this one mathematica.stackexchange.com/questions/5817/… are all different but the underlying issue is the same. I interpret your answer as they should be answered separately, correct? – Kuba Oct 8 '19 at 11:06
  • I don't know - how can you search for questions based on the duplicate they were closed as? – Jason B. Oct 8 '19 at 15:13
  • 1
    Duplicates are on top of the "Linked" section on the right side of the content. – Kuba Oct 8 '19 at 17:01

You must log in to answer this question.

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