Notices by Steve Canon (steve@discuss.systems)

@cr1901 @pkhuong @mcc there’s just a special fast-path for them

@cr1901 @joe @pkhuong dependency-breaking xor/movzx are "free" (they have to be decoded, but they don't need to execute).

@joe @cr1901 @pkhuong Intel's dedication to making that particular mistake _over and over again_ will never cease to amaze me.

@joe @cr1901 @pkhuong (They did it for 8->16, 16->32, and SSE->AVX/AVX512. And each time they managed to find a way to make it worse.)

@ryanc it's the thing being licensed, but the device in question does not handle any text input or output (beyond five single-digit displays), so my question is really "why on earth is there a UTF8 decoder at all"?

@ryanc that is very much not the question

Looking up technical information on an induction stove we might use for our house, notice a link "Information regarding Open Source Software". It's all the collected OS licenses that the stove uses. First entry: "Flexible and Economical UTF-8 Decoder"

... what?

@dalias yes, I’m exaggerating, it is _possible_ to write conforming C programs. But still we live in a world where only a very small fraction of the C code written actually is strictly conforming.

@dalias C and C++ solve this by littering the spec with so many landmines that no one can write a “correct to the spec” program. Newer languages solve it by simply not having a spec.

Fortran and COBOL remain undefeated as the only widely-used serious languages.

@mekkaokereke we kicked the pie on Friday, then Jen asked me to make another Saturday morning. We’ll be on pie #3 soon.

@Catfish_Man My code is perfect, David.

@dalias @cr1901 @kenshirriff that doesn’t mean you can’t reduce the argument! log and exp aren’t periodic either (at least, not on the real line), and we do argument reduction for them as well.

@dalias @cr1901 @kenshirriff expanding a bit more, I would say argument reduction is any process you use to go from evaluating a function on a larger interval to evaluating a related function on a smaller (in the sense of “computationally easier”) interval.

@dalias @cr1901 @kenshirriff what they said. A few steps of CORDIC still useful for atan(2) argument reduction, more if the only fast operation you have is addition. Otherwise use polynomial approximations everywhere.

If you’re in or passing through Boston, the O’Keefe/Moore exhibit at the MFA is really pretty delightful.

Anecdata: this morning was the most crowded I've ever seen my polling place, and it isn't even close, including a line for new registrations out of the room and down the hall.

@fay59 I swear sometimes I think that I could get absolutely any nonsense API through evolution just by giving it a ridiculous enough name. Everyone will just focus on that one thing. =)

@inthehands No. partially because it’s unclear what people want to happen when the modulus is negative, partially because there’s no compatible division operation. There is one on Numerics main, though, and it’s easy enough to roll your own, as you show.

@fay59 Numerics has BinaryIntegers.shifted(rightBy: count, rounding: rule) fwiw.

@jrose @markmalstrom (which makes it even funnier because NSArray is yet another, different thing.)