We recently closed a maths usage question, Can I use both of setbuilder notations in one article?, which asked for whether it is acceptable to use two different comprehension notations in one article, on the grounds that it is off-topic. It is quite true that "a mathematical style question", as Ryan Reich put it, is not about TeX and Friends.

However, the question was clearly phrased in terms of TeX markup, and the semantics friendly approach to TeX can influence how one best answers these questions. Consider, hypothetically - what if there was a comprehensions.sty package out there with distinct comprehensions \pointcomprehension, \funspacecomprehension, with a configuration command that allowed you to issue such commands as (excuse the ConTeXtism) \setupcomprehension[point][separator=:], together with the configurations used in several textbooks?

In my opinion, the possibility either that such specialist support might exist for such usage questions, or might be desirable to have, is a reason for such usage questions to be considered on-topic, provided they have such an overt TeX&co. context. I consider that a usage question with such TeX-specific context is a different question, in that it asks for different kinds of answers, than one without.

  • 2
    Good point. But could you please spell TeX as "TeX" not as "Tex". Commented May 3, 2013 at 9:06
  • 1
    @Martin: "Tex" and "TeX" are the same spelling but different capitalisation. I use regular proper noun capitalisation except where there is special reason to use "brand" capitalisation. Commented May 3, 2013 at 9:08
  • 7
    @MartinScharrer - I've generally taken an anti-brand capitalisation stance, since I think it reduces readability and promotes an "insider/outsider" attitude I don't like. Plus the Tex community has far too many, too tricky brand capitalisations (pdfTeX, AUCTeX, ConTeXt, SciTE, TikZ; there's no consensus on the brand capitalisation of Metapost) - I can't remember half of them without looking them up and, as the UK Tex FAQ says, they are often silly. So I'd be reluctant to change my long-standing practice. Commented May 3, 2013 at 9:20
  • 12
    i hate to get into the argument about "brand" capitalization, but i was the person on whose desk the letter from honeywell's lawyers landed, requesting that knuth and ams cease and desist from using "TEX" because for honeywell, it was a registered trademark. (see chapter 1 of the TeXbook for a brief discussion of the fallout.) while i freely admit my own idiosyncrasies in the application of this principle in this forum, if i'm writing for print, i'm careful to properly recognize all trademark claims, and encourage others to do likewise. (even though i agree there are too many.) Commented May 3, 2013 at 14:31

2 Answers 2


As I see this question it's a perfect candidate for http://math.stackexchange.com (as already suggested by a comment to that question). They users there aren't bad in TeX either, because they use it with MathJax.

Because there is already a specialized SX site for this I agree that we should consider such questions as off-topic on TeX.SX. However, migration to Math.SX should be prefered before closing it as off-topic.

  • We have no migration path to Math.SX, though - we'd have to ask the qner to do that manually. Commented May 3, 2013 at 9:21
  • 2
    @CharlesStewart: A migration path would indeed be a good thing. However, moderators can migrate to any SX site. Commented May 3, 2013 at 12:01
  • 2
    I'm convinced that this can be made to work for questions involving maths notation, especially after noticing that Math.SX includes in its scope "Software that mathematicians use", not to mention the fact that Willie Wong is a moderator there. I'll enquire in meta.math.sx. (I remember submitting more-or-less this comment last night, but it seems to have gone astray). Commented May 4, 2013 at 7:58
  • Posted: meta.math.stackexchange.com/questions/9345/… Commented May 4, 2013 at 8:09

I agree that the distinction between the present question (roughly, "Which of two acceptable mathematical notations should I use?") and the hypothetical one ("Help me with this package that produces that notation") is a little fine, but one objection is that the question as written is explicitly not about help using TeX at all: the same question could be asked about Word's equation editor, or handwriting for that matter. It's purely a stylistic question.

Now, the answers that were given before closure actually do (some of them) address a TeX means of dealing with the underlying issue, but on-topic answers don't make the question on-topic, just as novel answers don't make a duplicate question unique (if I recall this is a current topic too). This aspect of the nature of the question makes it something of an XY problem, with the added twist that Y is off-topic.

Another connection is with the similar problem with MathJaX, where several answers agree that at least a significant part of that genre of question is off-topic, and moreover is explicitly on-topic elsewhere. MathJaX questions are arguable in the same way, in that even installation and troubleshooting questions (as discussed by Martin Scharrer's answer) are analogous to such questions about installing and troubleshooting a TeX distribution that are on topic here...but when applied to MathJaX they are not, because the tool is fundamentally not TeX even if it accepts some TeX code. And likewise here, the mathematical style question is fundamentally not TeX, even if it admits some TeX solutions (that may not even be appropriate! What if it's a journal or university style?).

I think the usual thing that's done for questions like this is to ask the q'ner to refine their post to the point that it is on-topic, specific, and answerable to their satisfaction. However, what if their question really is "should I use two notations for sets?"; in that case, I think that it really belongs on a math forum because it's really not a TeX question, even if by virtue of being a typographic language TeX sometimes has opinions about matters of style.

  • This isn't quite addressing the question I asked. I agree that maths usage is not in scope for this site, but it is possible that certain kinds of questions that are on their face asking for maths usage answers, are really served best by answers that are primarily about Latex. Commented May 4, 2013 at 7:57
  • I think I addressed that a bit: first of all, if that's so, then the question needs to be rewritten so it's asking for what it gets (since what it originally asks is off-topic). And second, what if the style question really best has a style answer? Note, the question was not "how can I make this style in LaTeX", which is fine, but "can I use this style in my paper?", which even if it has a great and clever macro solution, is still not really on topic.
    – Ryan Reich
    Commented May 4, 2013 at 17:04
  • @CharlesStewart and sometimes questions that are on their face asking a TeX question are really served best by answers that are primarily about math usage. For example, see tex.stackexchange.com/questions/103413/…. Commented May 8, 2013 at 21:34

You must log in to answer this question.

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