No, HH uses windows.h. @molly1935 does not, and also has a much higher LOC.
-
-
Replying to @cmuratori @pervognsen
It's 60k lines all inclusive and compiles in around 0.8 seconds on my i7 laptop.
1 reply 1 retweet 6 likes -
-
Replying to @pervognsen @cmuratori
how'd you do transcendentals? lookup table + taylor series?
2 replies 0 retweets 0 likes -
Replying to @sssmcgrath @cmuratori
P.S. Never use truncated Taylor series for approximation. Minimax approximation with range reduction.
1 reply 0 retweets 1 like -
Semi-exception: A few Taylor series have very fast convergence like exp(x) = sum x^n/n! for x >= 0.
1 reply 0 retweets 0 likes -
Even there you need argument reduction for negative x: exp(-x) = 1/exp(x). Try computing exp(-1000000) directly.
1 reply 0 retweets 0 likes -
Better yet: Don't roll your own math functions if you have any way to avoid it. :)
1 reply 0 retweets 0 likes -
As always, MUSL is a good place from which to cherrypick: https://github.com/esmil/musl/tree/master/src/math …
3 replies 0 retweets 8 likes -
Looks like their exp is a minimax polynomial (with range reduction as always): https://github.com/esmil/musl/blob/master/src/math/exp.c …
1 reply 0 retweets 0 likes
I feel like there is too much math in the math routines. They should make math routines with less math!!
-
-
it all boils down to a bunch of NAND instructions and those are really simple so no big deal.
1 reply 0 retweets 0 likes -
Replying to @jackmottTX @Sol_Escape and
it's really just the difference of a volt or two, no biggie
0 replies 0 retweets 1 like
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.