3 ms·
We aim to be better than Word—have you noticed anything that's better in Word than in MathQuill?
by laughinghan 11y ago
We aim to be better than Word—have you noticed anything that's better in Word than in MathQuill?
- ygra 11y agoA few convenience auto-corrects, like typing +- to get ±, or -> to get → (and => to get ⇒). While \pm isn't too hard to type or memorize, the arrows in LaTeX are horribly painful (IMHO). \rightarrow for the normal one, okay. But the double? \Rightarrow? \rightArrow? \rightdoublearrow? Argh! You seem to be doing that for a few other things already, e.g. >= changes into ≥. I actually like Word's way of getting upright text (and thus without the math spacing/kerning) by using double quotes around something better than using \mathrm. That may well be subjective, though (but it is shorter). On the note of double-quotes: I am using US International, which has quite a few dead keys: ', ", ^, ~. Usually (on Windows, at least), you get the bare character (without combining it with anything) by typing "+space, which doesn't work in MathQuill, it simply swallows it. The same issue may well happen for a subset of those characters in other European keyboard layouts. Especially ' (used for primes) is a common dead key. In any case, MathQuill seems amazing and a good step forward in my eyes. I know my way around LaTeX, but managed to take notes in math lectures with Word faster and with fewer errors, simply due to the fact that (a) the language is much nicer to type (especially with non-US keyboard layouts due to needing far fewer braces and backslashes), and (b) you immediately see what you get. Both things I find handled considerably better in Word than in LaTeX (Lyx might be an option for (b), granted). So thank you for creating MathQuill and making it awesome (instead of just delegating to MathJax on every keystroke ;-)) Generally, Word's math markup is described in UTN #28: http://unicode.org/notes/tn28/ http://unicode.org/notes/tn28/ so it's for the most part openly documented, along with reasons why certain defaults or behaviours were chosen.