15

There was this question where several (and clever) answers were given. But afterwards the OP clarified "I am using Mathematica 7 and these options don't seem to work", making most of them not valid (for the OP's particular need). A moment ago, this question was asked; the first answer was to use Callout, and again afterwards the OP clarified "My Mathematica version is 10. With this code I do not have any output".

Situations like this are confusing. I think there should be an obligation to provide the version the OP is using in order to not give answers that will be useless for the OP's particular problem, and to focus attention of the answerers on the true issue the OP is facing.

11
  • 3
    When people aren't using the latest version, they should indicate that. But how can you get them to do it?
    – Szabolcs
    Nov 30, 2016 at 14:04
  • @Szabolcs Judging from your comment, I guess it would be impossible to have an additional window, like the one with tags, to be filled in with the version... Or to somehow make a tag with version to be compulsory...
    – corey979
    Nov 30, 2016 at 14:14
  • 1
    That would require a change to the stackexchange engine ... there's no chance that will be added for the sake of a single site. Also, in the past there were discussions about the overuse of version tags. As I remember most agreed that it's not good to add version tags unless the question is specific to a non-current version (i.e. the situation you describe).
    – Szabolcs
    Nov 30, 2016 at 14:16
  • 1
    I agree that not saying that you only have access to a 6+ yr old version is annoying, but I think it's one of the smaller annoyances (compared to people not describing the problem clearly or just dumping code), and I don't see how it can be fixed.
    – Szabolcs
    Nov 30, 2016 at 14:19
  • Inspired by this thread, maybe placing a text like "Please also indicate the version of Mathematica you are using" in the box "How to Ask" would be possible, and would reach at least some users? I don't know how stuff like this is handled here, I'm just throwing ideas.
    – corey979
    Nov 30, 2016 at 14:23
  • 3
    On the other hand, answers can help many other people with the same question, who probably don't have the same version restriction, so I think it's OK to get all the answers out there. OP's less likely to get their own problem solved this way, but that's what they get for moving the goalposts ¯_(ツ)_/¯
    – Chris K
    Dec 1, 2016 at 1:00
  • 1
    @ChrisK I knew this will finally be brought up. My point is that if you don't know the version, and you build your answer with the most recent functionalities, in a more or less complicated manner, you may think "that's a good job, I'm done with it". And then you are told that nothing applies. If you know the version in advance, you could either a) go straight to the point, b) give both the required and an updated version of the answer, or c) explicitly write that "this does not answer the particular question but might be of relevance for future readers". Everyone benefits then.
    – corey979
    Dec 1, 2016 at 1:08
  • 1
    @corey979 For sure. Maybe text could be added to the effect of "If you're not using the current version of Mathematica, please indicate the version you are using"
    – Chris K
    Dec 1, 2016 at 2:04
  • I think that it should be indicated but I do not see much hope for getting it enforced. In the comments people who are intending to answer should ask about the version is one way.
    – bobbym
    Dec 1, 2016 at 5:00
  • I thought the main purpose of tags was to help in finding questions, not to indicate to potential answers what the context of the question. I think most of the time version tags are unhelpful for this.
    – Michael E2
    Dec 12, 2016 at 11:50
  • And this fruitless discussion could also be avoided if I knew at the very beginning what version the OP is using; none of the info in the question indicated the problem is due to v11.
    – corey979
    Dec 13, 2016 at 23:34

2 Answers 2

Reset to default
9

I've just answered such a question. I think the status quo is fine, it serendipitously allows for variations on how to answer the question at the expense of the poster getting his specific answer (and "accepted" rep points for the answerer with the wrong version, I guess). I think this is more positive for individual question-answer quality in the long run and the poster can always clarify while he has the attention of the community.

Also, it doesn't seem trivial to enforce.

3

No, requiring specification of Mma version is not helpful.

I think indicating your Mma version as a footnote to your question is a good practice if you suspect it's relevant to your question, and maybe as a heads-up to let people know that some answers using newer functionality won't solve your particular issue. But it seems like most of the time it's superfluous.

On the other hand, specifying the version prominently when it actually has little or nothing to do with the issue covered by the question is distracting, and can give the (false) impression that the answers only apply to that version.

One of the things I dislike about Matlab is the high number of breaking changes and churning out "new" versions twice a year. In general I think the Mma developers are better about building consistently on the strengths of the language, so having the absolute newest version of Mma is not essential for doing productive work that can be shared with others.

You must log in to answer this question.

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