Closed zOadT closed 8 months ago
Good point! Ran 3x PR commit consecutively, 3x master commit consecutively and repeated this 3 times. Everything seems fine 👍
fix: use combine operator order from box2d
Running Benchmarks:
✓ planck.js (current)
Results:
| Name | avg ms/frame | 5th %ile | 95th %ile | Ratio |
| ------------------- | ------------ | ------------------ | ----------------- | ----- |
| planck.js (current) | 12.94 | 11.798041999340057 | 14.33016699925065 | 1.00 |
| planck.js (current) | 13.01 | 11.887500001117587 | 14.298832999542356 | 1.00 |
| planck.js (current) | 13.18 | 12.029790999367833 | 14.547833999618888 | 1.00 |
| planck.js (current) | 12.83 | 11.65612499974668 | 14.266707999631763 | 1.00 |
| planck.js (current) | 13.22 | 12.081416999921203 | 14.534709000959992 | 1.00 |
| planck.js (current) | 13.02 | 11.900624999776483 | 14.34375 | 1.00 |
| planck.js (current) | 13.22 | 12.049542000517249 | 14.55120899900794 | 1.00 |
| planck.js (current) | 13.13 | 11.960750000551343 | 14.503916000947356 | 1.00 |
| planck.js (current) | 12.83 | 11.659291999414563 | 14.188333000987768 | 1.00 |
fix setMassData
Running Benchmarks:
✓ planck.js (current)
Results:
| Name | avg ms/frame | 5th %ile | 95th %ile | Ratio |
| ------------------- | ------------ | ------------------ | ------------------ | ----- |
| planck.js (current) | 13.54 | 12.306792000308633 | 14.869791999459267 | 1.00 |
| planck.js (current) | 13.14 | 11.965250000357628 | 14.521833000704646 | 1.00 |
| planck.js (current) | 12.92 | 11.742500001564622 | 14.336292000487447 | 1.00 |
| planck.js (current) | 13.09 | 11.844208000227809 | 14.573958000168204 | 1.00 |
| planck.js (current) | 12.92 | 11.737958000972867 | 14.373540999367833 | 1.00 |
| planck.js (current) | 12.91 | 11.786875000223517 | 14.235790999606252 | 1.00 |
| planck.js (current) | 12.92 | 11.671540999785066 | 14.627749999985099 | 1.00 |
| planck.js (current) | 13.63 | 12.491916000843048 | 14.934125000610948 | 1.00 |
| planck.js (current) | 13.32 | 12.188791001215577 | 14.714042000472546 | 1.00 |
This PR is split in 2 commits:
Note: #255 also often made use of the (mathematically correct) assumption that
(A - B) . C == A . C - B . C
. Technically this also changes the order of additions, but I guess this is probably fine.