munrocket
munrocket
## length, distance Based on `inverseSqrt` internally, `sqrt` need polyfill here when length=0 or distance=0 > follow JS and WASM rules. WebAssembly not specifying any special function except `sqrt`. Ecmascript...
## normalize sqrt(0), https://github.com/gpuweb/gpuweb/issues/3220#issuecomment-1326802134 ## tan Defined as sin(x) / cos(y), division by zero if cos(y) = 0. ## smothstep If high = low, clamped division by zero is happening....
Thanks! It works after renaming. Just install .zip after this.
Probably my implementation is fastest on the market https://github.com/munrocket/double.js
You can't optimize proposed BigDecimal in base10 good enough, because hardware not support it yet. I am agree with @jakobkummerow by the way and https://github.com/littledan/proposal-bigdecimal/issues/31 Consider good sofware handles for...
Recently I am tried to implement modern arbitrary precision library CAMPARY in javascript. I have pretty good [early results](https://github.com/munsocket/jampary/blob/master/README.md) that can be probably improved x2 with better memory management and...
@Yaffle it was maid just for fun and based on Campary which is truly arbitrary. I can’t guarantee full precision right now and it have a room for performance improvement....
Interesting papers: * Implementing decimal floating-point arithmetic through binary * A Software Implementation of the IEEE 754R Decimal Floating-Point Arithmetic Using the Binary Encoding Format Don’t know why nobody posted...
Started to love this feature. Constructors vecN() very useful when you porting shaders from GLSL.
Computers don’t work in base10 arithmetic and programmers don’t care about non technicians. If you will do any computation in base10, libraries like three.js will be slow.