My wife is a math teacher, and the piss-poor experience of writing mathematical notation outside of MS Word keeps them stuck on Word. As much as she and I love LaTeX's math notation, she can't get her departmentmates onboard with that kind of syntax.
They need to maintain notes, guides, tests, quizzes, etc. and basically running a team OneDrive is really the only option because all the alternatives utterly fail a group of non-technical mathies.
Microsoft Office has two internal math formats, one of them is Ecma Math[0], the other the other is the "Unicode Nearly Plain-Text Encoding of Mathematics"[1], it is a Unicode standard and uses only Unicode standard characters. It has the unfortunate property of being hard to type on its own (the integral character isn't on most people's keyboards), but it is pretty easy to read as just plain text. If you copy an equation out of Word you'll get something like this: ∫(x^2/2)
[0] https://blogs.msdn.microsoft.com/murrays/2006/10/06/mathml-a...
[1] http://unicode.org/notes/tn28/UTN28-PlainTextMath-v3.pdf
Disclaimer: I work @Microsoft improving on some math features, and we are the main implementors of the spec. Which makes me sad, since it is an open spec and it is really powerful!
Maybe it could be integrated into the Unicode shaper (HarfBuzz/Uniscribe/AAT) with a language code for "maths".
Unfortunately not! I am wondering if I should make requests to the Windows team internally, the Managed wrapper class they ship has the math features compiled out. :( I'm in Office, where we use an internal UWP safe C++ version of Rich Edit. We bundle it with our AppX and load it up like any other C++ DLL.
Please post over on https://wpdev.uservoice.com/ and get others to upvote! If internal and external asks line up, it becomes much easier to argue in favor of doing a feature.
OK, I went and looked at the standard and it seems like they're thought about this: they allow input as either ∫(x²/2) or ∫(x^2/2) but they apparently output as ∫(x^2/2) because it's more general and editable, like if you wanted to change the exponent to something that didn't have a predefined Unicode superscript.
Thus assume only + - and * / are unambiguous, everything else be explicit about ordering.
EG: ∫((x^2)/2)
Storytime. A couple years ago, I was in the midst of copyedits on a book with a bunch of math in it. The copyeditors were using a different version of Word.
When they sent the edits back to me, the math was gone. Completely.
Not only that, but when I tried to copy-paste the math in from a prior draft, the Word file refused to save.
Eventually, I had to reconstruct every one of the damn things, by hand.
(Happy consequence: I caught and corrected an error doing so. But still!)
That would never happen in a plain text format. And this was the experience that made me abandon word processors for good.
LaTeX:
\begin{bmatrix}
0 & 1 \\
1 & 0 \\
\end{bmatrix}
[[0,1], [1,0]]
# this is a header
this is a header
================
the same could be done with something like ASCIIMath, where a plain-Unicode representation could be an intermediate form between ASCII and MathML. Why? Because keyboards don't type in Unicode, but you don't want to be storing only the final output - storing an intermediate Unicode form seems best, assuming you can keep modifyign it using ASCII and then going ASCII + Unicode => Unicode.
We also published a short post on 'the stoic resilience of the PDF within the digital ecosystem' recently [3], which seems relevant...although it's just a short background piece.
Mind blown. I only write a little math, and I can't imagine my first choice not being MathJax/LaTex in a plain text file.