This question is motivated by a series of questions in the past few days from one user who was clearly progressing through setting up a calculation and working through their errors one by one. Some of their questions in this series were troubleshooting questions, and some easily answered with a link to documentation. Such questions are not inherently bad, for example, python stack overflow is full of them and is often my go-to for my own python quick googles. But they are more likely to be written in the spur of the moment and, I suspect, more likely to be abandoned even if answered well, without the answer being accepted.

So I would like to broaden the discussion towards what can be good practices or advice we can recommend in the context of making calculations work before starting a question. The "homework" etiquette of chemistry.se is similar, but I cannot seem to find a citable formulation of it. Examples I can think of:

  • Check documentation, especially if it is good. For example, LAMMPS and VASP have decent or better documentation, arguably better than basic python.
  • Try a couple of things to solve the problem, even if you do not know what can help.
  • Consider what is fundamental to the question and what is not. We seem to see a lot of questions more or less "how do we do XYZ for system ABC" where the specific ABCD could plausibly be exchanged for something else. The specification of the system ABCD makes the question less likely to be found when another user later needs to do XYZ for system EFGH.

In my ideal world, I would be able to point back to this discussion when commenting on questions that fall in this category. I will consider a separate discussion of didactic or tutorial-type answers helping folks troubleshoot rather than providing light-on-context standalone answers. In the simplest terms, this one is for questions answerable by "it is in the documentation".


  • 1
    $\begingroup$ +1. Thanks creating this thread. In order not to dominate the Meta discussions, I'll let others provide their feedback before me! $\endgroup$ Commented Apr 27 at 21:05
  • $\begingroup$ In HW-like questions, users tend to delete their question in order to not leave any virtual footprint. I am not sure if this is the case with this deleted question, but it feels like one. The user complained that the answer provided was not adequate but did not mention what was missing. That's why I suspect this might be one of those cases. $\endgroup$ Commented May 2 at 17:11
  • 1
    $\begingroup$ I contend that the particular question you linked was (1) not a question solved by documentation. The physics in these diagrams is legitimately not straightforward, and IMO part of the physical meaning remains missing from the already given answers. (2) We as answer-ers and editors did a rather poor job of helping the user feel welcome. We (both of us, you and me) told this person far too many times that they were wrong instead of letting them work through it and arrive at better wordings and a better understanding of the physics. It's not trivial to ask a well-worded question! $\endgroup$ Commented May 2 at 21:12
  • $\begingroup$ I understand. My point was if it was suitable to include this criterion for HW-like question where user deletes the post after receiving an answer. However, this particular post is a bad example as the initial edit by me made the OP angry and I should have sought de-escalation instead of defending my action. From now on, I will not edit questions directly but I remain somewhat conflicted about the MMSE policy on more than one question per post. On a separate note, if you want to write an answer, I will ask a question about the physical meaning in DFE diagram. It would be really helpful. $\endgroup$ Commented May 3 at 8:30


You must log in to answer this question.

Browse other questions tagged .