On one hand we have discouraged writing "answers" in comments, and on the other hand, recently some answers were converted to comments (which "♦ moderators" are able to do).

We do want to encourage people to write answers, and we do want our (# of answers) / (# of questions) ratio to go up, but at the same time we do want answers (which allow users to gain rep and badges) to have some minimum standards. What do we want those minimum standards to be?

I'm sure there's Meta posts all over the network that say that just linking to a paper ought to be a comment rather than an answer, and that one needs to actually explain the relevant contents of the paper in a self-contained answer rather than just posting a link to a paper and making people read through the paper until they (may or may not) find a part that's relevant to the question.


2 Answers 2


For me, the answer to your question is simple: I would like to have answers that really answer the question, even partially (sometimes there is no space or time to write/give a full answer).

Telling someone that what he/she is asking is in "any textbook" is not an answer as telling to Google it isn't too (even when could be true).

  • $\begingroup$ Agreed, this is also the case for e.g. software SE. A link to the documentation isn't an answer. An answer is self-contained and if content is taking from articles, they should be in quotes and clearly specify from where. This however, is probably a problem due to copyright issues of publishers. So the quotation would only work under articles @ arxiv, or open-access publications. $\endgroup$
    – nickpapior
    Commented Aug 19, 2020 at 10:35

I don't understand why my answer to

What functions other than Gaussians are used for orbital basis sets?

was deleted and converted to a comment. I answered the question, and the OP was happy about it. Comments can't be accepted as answers, and remain only visible for some days. (Maybe not all the moderators are aware of this?) This kind of behavior really discourages answering straightforward questions, especially in this kind of a case where there's an in-depth answer available in an open access article.

There's nothing wrong with answers that are short and straight to the point.

  • $\begingroup$ I think the communication wasn't great on our part in addressing your answer, so I'm going to undelete it. However, I probably would have left a comment initially asking you to add a bit more context to the answer explaining some specifics/highlights of the linked papers. While its fine (even encouraged!) to include links to supporting material, SE wants to avoid "link-only" answers $\endgroup$
    – Tyberius Mod
    Commented Aug 7, 2020 at 21:47
  • 4
    $\begingroup$ I really don't think that an answer that point to a paper/manuscript should be considered as a good answer. The idea of any SE site is that it should be self-consistent: the question/answer forming a closed system. If I make a question I would like to have, at least, some sort of answer that solve my doubt even partially. Of course, any reference, bibliography that could guide me in looking for a full (or better) answer is welcome. $\endgroup$
    – Camps Mod
    Commented Aug 7, 2020 at 22:34
  • $\begingroup$ OK, I've expanded on the answer. $\endgroup$ Commented Aug 8, 2020 at 15:25
  • 2
    $\begingroup$ @Camps I guess there should be an exception to that rule for questions tagged reference-request? In that case (and depending on the question), pointing to the paper can fully answer the question. $\endgroup$
    – Anyon
    Commented Aug 8, 2020 at 18:08
  • $\begingroup$ @Anyon Interesting point. Though everything with the reference-request tag so far: mattermodeling.stackexchange.com/questions/tagged/…, that has received an answer, seems to still have quite a lot of substance in the answer: I haven't found any that would fit in a comment, for example (though I did not look through all 50 questions with that tag). $\endgroup$ Commented Aug 11, 2020 at 1:42

You must log in to answer this question.

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