18

We often have the case where two questions are asked by users facing completely different phenomenon, but where if you have the right understanding, you can see that the same thing is going on in each case. In his answer to the meta.tex.sx question Different questions with the same answer, Andrew argues for merging these questions. This is good, provided it is clear to people seeing either of the symptoms that the answers address their problem.

However, in practice these questions are often getting closed as exact duplicates of each other. From the experts point of view, this may seem natural, but a user lacking this expert knowledge will see that instead of being given a solution to the symptoms they see, they are being told how to treat a completely different set of symptoms. Without an explanation of why the two phenomena are about the same problem, this is a failure to answer the question.

I propose that our policy in this case is either (i) merge per Andrew's suggestion, ensuring that the question makes clear that answers should address both sets of symptoms, or (ii) tolerate different questions with analogous answers, provided that links between them, and our policy is also that we do not (*iii) close questions about one set of symptoms as being exact duplicate of another question that deals with quite different symptoms.

This question is in response to the closure of Change abstractname.

3
  • 1
    I added a (hopefully) comprehensive "naming" Q+A: tex.stackexchange.com/questions/82993/…
    – lockstep
    Nov 16, 2012 at 14:30
  • @lockstep - this is excellent, thanks. Both could be closed as duplicates of that question, but it might make sense to have a two-line answer for the abstractname qn pointing to the part of the answer that is relevant (e.g. "Provided you aren't using the babel package you can \newcommand{\abstractname}{...}, see qn 82993"). Nov 17, 2012 at 9:25
  • Actally the OP commented that the problem was resolved by setting up babel correctly. I added a link to the "canonical" Q/A at the start of the OP's question.
    – lockstep
    Nov 17, 2012 at 11:21

4 Answers 4

13

I would have to disagree here with Stefan here.

While I am not what most would consider a newbie, I am by no means an expert on general LaTeX usage as I have mostly focused on only what I need at the moment, or think I will need coming up in the foreseeable future. And hence, I don't use abstract, or figure names.

Looking at those two posts, they seem totally unrelated. I have done enough hacking with TeX that I think I could get figure this out. But by just looking at the answer in the duplicate, it is certainly not at obvious to me that these two are related, or even where I should begin to look for the problem. If I was a new user and had a question about changing the abstract name and it got closed as a duplicate of "Change caption of names of figures" I would really be confused, and discouraged.

While the OP did not provide a complete MWE, the relevant code snippets and document class were provided. It seems like a decent question. Since the problems appear so different, what I would prefer is to have the answer that solves the problem at hand and a cross reference to the other question. Only then would it be clear as to how they are related.

I also dislike the idea of changing the original question to a more general one. I constantly refer to many of my older questions and if it were changed significantly (either title, or body), words that I remember being part of the question might no longer be in the text and would make it harder for me to locate (Search on TeX.SE is already very difficult). And changing the title from "name of figures" to "how to change names of objects" is a perfect example of such potential confusion. I would be thinking I had a question about "figure" and not "objects". Sure, after careful thought one would come up with "how to change names of figures, abstracts, and other objects", but not sure that people changing the question would always see that.

I realize that the site is not just for me, but if I can't access my own questions, I would end up asking even more questions -- I am sure no one wants that!! :-)


Ok, just saw Mazzy's comment after I looked one more time to make sure that his post has some merit -- I guess that is the clue of where the problem was.

And closing as a duplicate to the one lockstep linked to above would not be so bad. Then the other older questions should be cross referenced to this new question.

11

I think that another point is missing here: Answers for caption are different than answers for abstract. There is a common thing -- changing setup of \...name, but for captions there are other possibilities, and this makes it hard to "read only the part of the other question that is relevant to me, when this part is not really seperated". So in my opinion, this is even not the same disease, only a similar one.

And one more point: When you vote for "exact duplicate", the two things mentioned are "same content" and "identical question", and that is not the case here. On one hand keeping many "too localized" questions that could be closed by saying: "it's your problem, not LaTeX's one, and it's nearly impossible that someone else will encounter the same problem". And on the other hand, we close questions that are maybe very similar, but still very different? No, I don't think this is the correct approach. Having similar questions does not hurt that much in my opinion.

1
  • 8
    posting a comment that "question x is related, but not identical" could be an appropriate response. Nov 17, 2012 at 13:23
6

(i) If a question is closed as a duplicate of another one, it's good to check the other one if it can be rewritten to be more general, so it's obvious that it applies to the other case too. In this example, the "name of figures" question can be changed to "how to change names of objects" or the like.

(ii) There's no need to tolerate different questions with analogous answers. Just for the OP to feel better? To save a click for future visitors, collect a huge amount of pretty identical answers, spread across differently worded questions or specific cases? The future visitor, even coming via a search engine to the specific case, is helped by the prominent duplicate link to the more general solution. So closing is good, if the other question and answers solve the problem. Note, each duplicate link makes the linked question more prominent and also higher in the FAQ list.

It's a good question, Charles, I don't mean you but as a general idea: if something such as closing as a duplicate doesn't seen to fit perfectly, instead of denying closing it would be good if you firstly check if it could be fixed by improving the other question. Making questions more generic and canonical at such an occasion is great.

2
  • Doing a proper job of merging is the best case. But some of these "exact duplicates" will look like nonsequiturs to Tex* newbies. Nov 16, 2012 at 12:18
  • 2
    Can be fixed by a comment or an edit. Merging answers can be a problem: I would not merge specific-case-answers to the generic question which already has a generic answer.
    – Stefan Kottwitz Mod
    Nov 16, 2012 at 12:21
2

Shouldn’t this question been closed as an exact duplicate of this question? =:-J

In my opinion, two different question with the same answer is not exact duplicates. The second question may be answered by referring to the first one, but not closed as a duplicate. Because it is not a duplicate!

You must log in to answer this question.

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