Of course, for complicated formulae TeX syntax is unavoidable. I have in mind some simpler situations such as

  • 2.3=6, 2*3=6 or 2x3=6 vs. $2.3=6$, $2\cdot 3 = 6$ or $2\times 3=6$
  • for any choice of a and b vs. for any choice of $a$ and $b$

Of course the Mathjax output looks better, but I found both of them readable. Perhaps some users with slow Internet connection can experience problems with too many formulae in the text. (I have only experienced problems when editing some posts that were long enough, but switching of the preview is a simple solution to this. Occasionally IE9 is unable to display some answers, but Firefox works fine.)

Is it advised to use Mathjax even for alone-standing variables (as opposed to italics)? Is it advised to use Mathjax for simple computations, like addition or multiplication a few numbers? (Probably you can think of more similar situations.)

Do we have official policy/general advice on this?

At wikipedia, it seems that for simple formulae they avoid latex markup: http://en.wikipedia.org/wiki/Wikipedia:Manual_of_Style_(mathematics)#Very_simple_formulae

I have tried to search for similar questions, my apologies, if I missed something.

  • 2
    $\begingroup$ your first example looks like 2.3=6 to me.. $\endgroup$ Aug 6, 2011 at 8:29
  • 19
    $\begingroup$ One objection to your first formula 2.3=6 is that some people are not used to periods as multiplication signs and may be confused by such an equation. Also, I don't like the idea of mingling different fonts *x* is x while $x$ is $x$ and they are set differently. The difference is minor but I find that distracting. My main objection though is that I don't like official policies on such trivial matters. $\endgroup$
    – t.b.
    Aug 6, 2011 at 8:36
  • $\begingroup$ @Martin: The first objection Theo raises is in fact the reason I edited your answer. Although there may be an argument for not using the centre-dot either (since, apparently, that is used as a decimal separator too). I was thinking about changing it to $\times$ but $\cdot$ seemed more faithful to your original post. $\endgroup$
    – Zhen Lin
    Aug 6, 2011 at 10:33
  • $\begingroup$ @Zhen: I have nothing against your edits. (If I did, I would post a comment under the answer you edited.) In fact, I am grateful to anyone who tries to improve answers, which helps people using this site. However, your edit reminded me that, on several occasions, I tried to avoid Mathjax using italics for the reasons I mention; so I wanted to know what opinion other users have on this matter. $\endgroup$ Aug 6, 2011 at 10:38
  • 1
    $\begingroup$ @Martin: In quite the positive probability someone would have edited the answer into $\TeX$ format anyway... $\endgroup$
    – Asaf Karagila Mod
    Aug 6, 2011 at 13:57
  • 2
    $\begingroup$ If we were discouraged from using TeX, answering questions in readable form would be too much of an effort. $\endgroup$ Dec 3, 2011 at 16:50
  • $\begingroup$ Wkipedia converts simple math markup to plaintext, and complicated markup to PNG. So it's not related to the lag the user may experience, it's related to server load. $\endgroup$ Mar 21, 2012 at 4:25

1 Answer 1


Well already there is a problem, to me it looks like you just said


two point three is equal to six

Which is .. crazy :)

I think it's best to just use MathJax in most places, even for trivial math, since it is the most specific and clear way of rendering math that we have.. the initial performance penalty of the MathJax dependency is already paid per-page anyway due to the nature of how the web works.

Also, web browsers and computers get faster every day, and we will be contributing some performance fixes back to the MathJax project as a MathJax partner... there's lots of room for .. uh, shall we say, improvement, there. :)

  • 3
    $\begingroup$ Jeff: Assuming $0=1$ then $\frac{23}{10}=6$. :-) $\endgroup$
    – Asaf Karagila Mod
    Aug 6, 2011 at 9:28
  • 21
    $\begingroup$ @Jeff Thanks to you and the SE team for the great MathJax support. No doubt it plays a nontrivial role in the success of Math.SE. $\endgroup$ Aug 6, 2011 at 16:37
  • 5
    $\begingroup$ Personally, it takes me more effort to remember the Unicode for a centered dot than to just type $\cdot$ within dollar signs... $\endgroup$ Aug 7, 2011 at 5:16
  • $\begingroup$ @Asaf : Assuming $0 = 1$ one can prove many things... such as $\mathbb R = \{ 0 \} = \{1 \}$. $\endgroup$ Aug 7, 2011 at 17:55
  • $\begingroup$ @AsafKaragila Would that be a good question ('What all can we prove if we are given that $0=1$') or will it just get closed? $\endgroup$ Feb 9, 2015 at 19:03
  • 2
    $\begingroup$ @ghosts: Way to pick a comment from 3.5 years ago. It would be a decent question, but the answer is nearly trivial. If you work with some "usual" axiomatic system from which we can normally prove $0\neq 1$, then by adding $0=1$ we prove a contradiction and then the principle of explosion (look it up) says we can prove everything; or we didn't assume anything that would allow us to prove $0\neq 1$ in which case you still can't prove that $0\neq 1$ by adding $0=1$. Finally, you might be using paraconsistent logics, or some other non-classical logic, but then you should know the answer yourself. $\endgroup$
    – Asaf Karagila Mod
    Feb 9, 2015 at 19:24
  • $\begingroup$ Jeff Atwood is the real wtf. $\endgroup$
    – nomen
    Dec 26, 2019 at 19:45

You must log in to answer this question.

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