'(Robert Smith)

> be NASA programmer > excited to write Lisp > code lookin like hieroglyphs > boss walks by > "wtf is that, alien code?" > no, just Lisp boss > boss laughs, Lisp now forbidden > mfw no cosmic parens > mission to Mars now in JavaScript > planet doomed > feelsbadman.jpg

Standard tech company interviews filter these people out by construction.

It's even sometimes a bit of a drag to have a conversation with these people on something new. Lots of mumbling "iunno..." Bohr-smart because Niels Bohr was allegedly like this. It defies our usual imagining of what a smart person acts like.

I've met many smart people that are slow-smart or Bohr-smart. They often don't have a quick answer, they struggle to contribute in meetings (esp. with loudmouths present), but give them a day or week, and they come back with the most brilliant, insightful thoughts and answers.

@reggieperry If the GC isn't parallel then whenever a GC happens, all threads are stopped.

@stylewarning I’m not up on SBCL internals. Will parallelization be hurt if the garbage collector isn’t parallel/concurrent?

OpenMP-style loop parallelization annotations in Coalton?

@reggieperry I mean it would be direct syntax, but the principle is the same: give some guidance on how you want a loop to be dealt with

@stylewarning Why use annotations instead of direct syntax? Those other guys need pragmas and suchlike because of the limitations of their languages.

@zenck_null @inerati Every 10 is 10 in base 10.

@stylewarning @inerati Two?

why is all software so horrifically bad

I always thought HOAs and their boards were being caricatured but, no, it's actually really how it is. Frivolous hearings, "emergency" assessments, "vote or no keys to the pool", etc. etc.

@furiouswhopper @run_guns @nicbarkeragain complete non-answer

@stylewarning @run_guns @nicbarkeragain Code should read like prose, or as close to it as possible. Any language incapable of that, or which flies in the face of that goal, is bad from the start.

Programmers: do you have a consistent preference for which side you put the literal on in a comparison? // Right-hand side if (someValue == 1) {} // Left-hand side if (1 == someValue) {}

@Frinklang Oh yeah, for the tremendous respect I have for GMP, I hate the interface.

@stylewarning @jamesladd I think that GMP's performance is what we should all aspire to! (You may have beaten it already LOL.) But definitely not its interface. I spent 11 years getting the Java people to let me make BigInteger and BigDecimal faster and I gave up LOL. Threads: x.com/aeliasen/statu…

10 billion digits achieved. 1 week of fixing very difficult numerical, multithreading, verification, memory/disk management bugs; modest hardware upgrades; and speed issues that were attacked by improving SBCL further. On to 100 billion.... if I have enough disk... and time...

@scheminglunatic it's a form of immersion in the art

@stylewarning making proramming language gjijnka / pl-tan yaoi is definitely a normal hobby for people to have ,right?

I've been writing C code like a Standard ML programmer and a FORTRAN programmer fighting for custody over the keyboard

@scheminglunatic wow a use for gemini

@stylewarning i got gemini to make me a saucier version of C++ / Rust FFI yaoi but idk if i should share it [again]

@Frinklang @jamesladd If you checkpoint the series it works, as it looks like you do. If you only have the digits, I wrote a program that does a nonic fix point convergence (given N correct digits, produce 9N correct digits). jury's still out of it can be made faster than simply recomputing chudnovsky

@stylewarning @jamesladd The more recommended version is this one which is more complicated but *resumes* calculations when you ask for more digits. If you calculate a million digits of pi, and then a million + 10 digits, it resumes the calculation efficiently. frinklang.org/fsp/colorize.f…

@scheminglunatic i want fast code that's not shitass C++ whilst larping as a functional programmer (left) who himself is larping as a functional programmer (right)

@stylewarning That is fascinating, I think. Please, tell me more.

@Frinklang @jamesladd very cool, didn't know this existed :D

@stylewarning @jamesladd Nice! Thanks for the implementation. Frink's "simple" implementation uses binary splitting which of course makes the implementation incomprehensible but is pretty readable as near-normal mathematical notation: frinklang.org/fsp/colorize.f…

@Frinklang @jamesladd it compiled fine for me ca. a week ago. and i'm happy we can read the source of GMP. many of the authors contributed great literature too, like Zimmermann's MCA: members.loria.fr/PZimmermann/mc… It can be terse mathematics and computer programming, but at least it's available

@stylewarning @jamesladd And then fix the compilation errors for a modern compiler LOL

@Frinklang @jamesladd mostly same algo in Coalton, but it's just as uncommented and cryptic 😔 #L554" target="_blank" rel="nofollow noopener">github.com/coalton-lang/c… at least we link a paper to learn about binary splitting :D

@stylewarning @jamesladd My friend, I have tried to understand that cryptic uncommented code for the last decade LOL

@Frinklang @jamesladd (alternative: compile gmp-chudnovsky.c in one short command on any platform and read the source code to learn about it: gmplib.org/pi-with-gmp)

@Frinklang @jamesladd (so long as you have a windows or linux intel machine and you're into running closed source binaries)