$ node
    > 0.1 + 0.2 - 0.3
    5.551115123125783e-17

Ruby, Perl5, Python also fail this. Perl6 uses Rat numbers for things like this and doesn't have this problem.

Do the same with a proper decimal datatype, and the same operations give the correct result - pay out 1/3=0.33; pay out 1/3=0.33; the last payment is 1-0.33-0.33 = 0.34 exactly.

The same goes with all kinds of other issues e.g. totalling up tax or discount amounts (calculated as a percentage) where you'll get mismatches if rounding happens at the wrong time, etc.

E.g. ten separate receipts each selling a 0.10 item with a 9% tax or discount will result in each receipt having 0.01 tax or discount; so the proper total in this case is 1.00 sales with 0.10 tax or discount; different than the case if you sell ten such items in a single transaction. If you do all the calculations in floats, you get this wrong.

  (1.005).toFixed(2) // Expected rounding to "1.01"
  "1.00"

for (let n = 0; n < 1; n += .01) { console.log(`n: ${n}, half: ${(n / 2).toFixed(2)}`); }