There isn't an option for that currently. But in your specific case, I think if you make box2d use doubles instead of floats, it should be identical, since the issue is only with floats.
On Thu, Jul 27, 2017 at 4:02 AM, caiiiycuk <[email protected]> wrote: > Hi. Is it possible to enable precies float computation for a part of code. > I use box2d and it works incorrect in PRECISE_F32=0 mode. For each frame > native version makes 1-5 physics steps, but js version makes 100-200 steps. > But if I set PRECISE_F32=1 then both versions works identically. I guess > that problem is in b2World::Solve function, maybe some corner cases do not > work as expected because of rounding errors. I want to compile only this > function/ or all box2d functions with PRECISE_F32=1, but rest of project > with PRECISE_F32=0. Is it possible? Or maybe I can easily rewrite C++ part > to avoid rounding errors? > > -- > You received this message because you are subscribed to the Google Groups > "emscripten-discuss" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/d/optout. > -- You received this message because you are subscribed to the Google Groups "emscripten-discuss" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
