# Is it OK to abuse Mathjax for emphasis?

The author of this question uses Mathjax (in a way different from its intended usage) to make parts of his question stand out more; for instance, to display some headers in red or to alter the spacing between paragraphs.

I believe this should be edited out, for various reasons:

• it is not proper Markdown syntax
• It breaks screen readers for visually impaired people
• It is ultimately a way to abuse the options offered by the syntax of this website to gain more visibility. If the developers decided that black, non-bold 19px Arial is the normal way to display headers on this website, should I be allowed to override this setting because I want my header to be in red so it stands out more?

I did edit it out, and the edit was reverted by the author because it "lets him get the presentation he wants". I would like to know what the community and the moderators think about this issue.

In other related questions Using MathJax for formating (**bold**, *italics*)? and On using MathJax to create 'white space', it seems like the leading answer mildly discourages this sort of abuse.

• If the author had applied \mathsf or even \text to make the typeface closer to the rest of the text it wouldn't be so bad... :-P but anything that can be done by markdown should never be done by MathJax. Getting colour like this is creative, and not otherwise obtainable, afaik, so I'm torn. Oct 18 '18 at 13:13
• Eg $\color{red}{\textsf{Question}}$ from \color{red}{\textsf{Question}} but the size is not right.. Oct 18 '18 at 13:23
• A website is better to have a homogeneous typography style. Oct 18 '18 at 14:29
• It may be worth mentioning that the post has now been edited to remove the offending MathJax. (I was a bit puzzled when I went to see what the fuss was about, until I looked at the edit history.) Oct 18 '18 at 16:08
• I had cause to do exactly the abuse in my previous comment, here: mathoverflow.net/questions/313527/… as there is a proof assistant whose correctly formatted name involves half of the letters coloured red. Oct 24 '18 at 21:44
• @DavidRoberts I don't like it either. I more or less agree with this article on this kind of stunts in logos and names. Oct 25 '18 at 7:56

I agree that text formatting should be accomplished with Markdown if at all possible. Since Markdown is used across Stack Exchange, there will generally be better support for proper rendering across browsers, mobile devices, screen readers and other accessibility tools, and better compatibility with future systems. MathJax, by its nature and because it is third party software, is going to be more subject to problems in these areas.

MathJax therefore should not be used for ordinary text formatting, unless there is a clear and specific need to create an effect for which there is no reasonable Markdown alternative. Simply varying the size or typeface of headers, enumeration numbers, etc, doesn't meet this criteria, and such uses of MathJax can and should be edited back into their Markdown equivalents. Wanting to have colored text is also not a good reason to do this; custom colored text in Markdown has been declined by Stack Exchange in general.

If a poster feels they have an essential reason for using MathJax, and it is edited out by someone else, they can roll it back and post a comment explaining their case. They can also put an invisible comment in their post <!-- like this --> which will only be seen by editors.

• "generally be better support for proper rendering across browsers, mobile devices, screen readers and other accessibility tools," <-- this. I find that on mobile Chrome the MathJax doesn't know how to space properly, so the text immediately following it is moved left so that it overlaps partially with the content within $...$. This is a pain, but even worse when an author does it just for text. Oct 18 '18 at 22:45
• @DavidRoberts If you encounter some problem with MathJax (like the one mentioned in your comment), it might be reasonable to post a bug report somewhere. Both Meta MathOverflow and Mathematics Meta might be an option - since they are visited by several people from MathJax team (Davide Cervone, Peter Krautzberger). It seems that they are quite active on Mathematics Meta - maybe a bit less on this meta. (But it might be just an impression caused by smaller sample.) Oct 19 '18 at 6:57
• No, it's a browser problem, as it happens on the nLab and the nForum as well. Oct 19 '18 at 10:28
• @DavidRoberts As this is only tangentially related, I do not want to leave too many comments about this (=reporting MathJax problem) here. If there is more to be said we can continue this discussion in chat. I have also asked a question about this on Mathematics Meta: Where to report MathJax bugs? Oct 19 '18 at 11:21

I think the question is too emotionally laden to present the problem correctly. To me it is like "All the preprints use Knuth's computer modern font. How dare we change that?". The use of the word 'abuse' seems to have already not only judged the action, but tried and convicted it, and sent a summons to the executioner.

A more charitable presentation would be along the lines of " This style really jars. I find it hard to read and distracting. Shouldn't we insist on uniform styling of all posts?"

I say not. That is what the vote counts and comment boxes are for. If the presentation causes dismay in the community, let the community speak. I don't like the style that much either, but I will not speak for the community on this. Best if I comment what problem I have with it, down vote it if needed, and move on. If the community ends up accepting the style, then I will probably avoid that author's questions (but not without giving my justification for avoidance).

Gerhard "Has Choice In His Reading" Paseman, 2018.10.18.

• Mathjax is a library to display mathematical notation. Using it to write text in red instead of black is "improper, incorrect, or excessive use; misuse", in my view. Do you have a less emotionally loaded word to suggest to indicate that this is not the usage Mathjax was meant for? Oct 18 '18 at 14:28
• Also, your pre-prints metaphor is interesting. I would argue that a more fitting version is that a post on my own blog is like a pre-print, where I am allowed to use the style that I prefer. But posting on this site is more like a journal article, where there is a prescribed house style and one is not allowed to write the title of their paper in a larger font so that it is more visible than the other articles in the same issue. Oct 18 '18 at 14:33
• @FedericoPoloni "one is not allowed to write the title of their paper in a larger font so that it is more visible than the other articles in the same issue." I didn't write my title larger, and I strongly object to the insinuation that I did. The formatting you object to is inside my answer, and is in no way analogous to the situation you describe since for anyone to see it they would have to already be reading my question. You seem committed to this picture of me trying to steal attention away from other questions; that's simply not true, and I'm rather put off by it. Oct 18 '18 at 15:12
• @Federico, I agree that there is an issue here, and I suspect it is social in nature. It is apparent to me that it is important to you, and I think it should be discussed. However, this is the fourth comment in this post, and I already see the beginning of the flame war. If both sides take a breath, apologize to one another, (or backpedal, or whatever), we can try again to find the problem and discuss it, with both emotion and consideration for the other side. Gerhard "Thus My Reframing Of Issue" Paseman, 2018.10.18. Oct 18 '18 at 15:34
• @NoahSchweber I see your point; I recognize that "writing the title larger" is an inappropriate metaphor, and I apologize for the insinuation and for my heated tone. I still think that a red text attracts more attention than a black one, though. If you wish, a more fitting metaphor is writing a section header in red. It only affects readers who are already casually browsing that page, I agree, but it is something that is there only to attract attention. Oct 18 '18 at 15:46
• @FedericoPoloni I still don't think I agree, but I certainly don't object as much. My goal actually was somewhat the opposite: I wanted to make the question easily skimmable, so someone could scroll through quickly and say "Oh, the question has [shape]," and then become more confident in their level of interest - either decide to read the question in detail, or realize it's not worth their time. This is why I don't think it was abusive in quite the way you're describing (and certainly adding a bounty is a more effective way of drawing attention). (cont'd) Oct 18 '18 at 15:49
• Regardless, your point about screen readers was completely surprising to me and obviously absolutely convincing: I didn't think at all about the problems my formatting would cause someone using a device of that kind. (That said, I think we do have a contentful disagreement over the extent of authorial freedom; for what it's worth, if not for the screen reader issue I think my response would have been to open a meta thread asking what community consensus was, and in the meantime side with the author, but I understand (I think) your position.) Oct 18 '18 at 15:50
• I also apologize for my own strong tone; I was caught off guard by this (although obviously that's not an excuse). I hope we're still on collegial terms. Oct 18 '18 at 15:53
• @NoahSchweber No strong feelings from my side. I'd be happy to offer you a round of (beverage of your choice) if we ever meet in person! Apologies again. Oct 18 '18 at 16:01
• Thank you gentlemen. It has become an honor to share a thread with you. Gerhard "Not To Mention A Community" Paseman, 2018.10.18. Oct 18 '18 at 16:09

Once typographers are allowed to run amok, there is no way back to normality. For it'd be only a question of time for multi-coloured flashing parts of text, and what not, to start to appear.

Contributions to MO are akin to papers in a journal or chapters in a book, where the editors are free to impose a uniform typography style.

Are there any Markdown MO deficiencies which need fixing, to provide more expressiveness etc? IMHO what we have is good enough.