3 ms·
FWIW, this would be a literal s-expression translation: (mrow (mi x) (mo =) (mfrac (mrow (mo -) (mi b) (mo ±) (s
by wtbob 11y ago
FWIW, this would be a literal s-expression translation:
(mrow (mi x)
(mo =)
(mfrac (mrow (mo -) (mi b) (mo ±)
(sqrt (mrow (mi b) (msup (mi 2)) (mo -) (mi 4ac))))
(mrow (mi 2a))))
And this would be a saner one, where mrow is implied:
((mi x) (mo =) (mfrac ((mo -) (mi b) (mo ±)
(sqrt (mi b) (msup (mi 2)) (mo -) (mi 4ac)))
(mi 2a)))
I think either of those is clearly and inarguably superior.