the early history of k5 is described in vector.org.uk/impending-kos, which is how most of us found out about it

I wrote oK because k4 was expensive, kona was based on k2, and I wanted to understand the latest developments in arthur's approach to programming as embodied in k5, but all that had been released to the public were a few cryptic examples on the kparc site

I reverse engineered as much as I could by studying the differences between k2 as described in its reference manual and those fragments

What is "special k"?

Are the games written under "games" at beyondloom.com written in K/

@ngn ffi?

the only games I have written in K are in the ike examples directory of the oK repository

most of my games are written in java, c, forth, or octo assembly language

forth or octo?

what happened to fith, sikth, senth?

it's best not to speak of those

they're hard to say anyway

snaKe is what snake looks like in K, with zero conditionals just for fun and to prove a point: raw.githubusercontent.com/JohnEarnest/ok/gh-pages/ike/examples/…

if it is "just to prove a point" then why wouldn't a stand along game project "prove the point more"

@ngn where are you finding your resources for optimized primitive implementation?

@nathanrogers ngn/k doesn't really do any super fancy primitive optimization. arith does have some unrolling to 256 bits so the autovectorizer kicks in, but otherwise it's just your average C

@nathanrogers you have kind of a persistent pattern of saying "well, why isn't your personal free-time project more elaborate???"

@dzaima (where by "average" I mean ignoring whitnification ofc)

Not really, I'm just saying if building snake in K proves the point that games can be made without branching in an array language, then why not use it as a tool for your other projects, rather than Java/C, etc

@dzaima that's a hard word to grok, maybe you could say whitneyfication?

@nathanrogers idk, it feels weird to me if the postfix isn't "-ification"

I don't in general use K to write video games because the available implementations are too slow, too license-encumbered, and/or insufficiently portable for practical distribution

@ngn ffi?

@nathanrogers if anything, I think repeating asking the same question will be detrimental to the cause

@dzaima can you ask ngn if the ffi is done?

ಠ_ಠ

is that a no?

well in lieu of that @JohnE what aspects of the languages are too slow for doing games?

@nathanrogers that's an "I hope your message was a joke"

Yes it was funny too

But the follow up question @u was not a joke, and is not funny

@nathanrogers well, the "ಠ_ಠ" also implied that I didn't really find it as such

joking about ongoing issues

I thought the emoji implied a humorous receipt

and also made me giggle

haha

I mean, I don't know the last time I asked. in srs brsnrs, he mentioned that I convinced him of the need for 2:, and I've checked in periodically to know whether its being worked on, close to done, implemented, so I don't see the harm in checking in every now and again

but 2h apart with no intermediate reply?

no, but as a humorous repetition in response to "too slow... insufficiently portable", as in suggesting that this would be a fix for that?

But I guess nobody pays attention to context anymore... that's too much brain stack at once

Can't remember the immediately preceeding message, but I can find matching repetitions!

"this guy is so dumb, he couldn't possible be making a joke about something, he's just annoying and stupid and repeats himself despite ngn not being logged in, lol what an idiot"

@nathanrogers using external libs would really make it less portable, and it wouldn't do anything about speed, unless you write your whole game in C and your k is just a call to 2:

as such, I interpreted the message as being unrelated to the surrounding context

in addition, the message started with an @ so I immediately read it as a reply (i'm used to concurrent discussions), the message had nothing to help take it as a joke (except being character-identical to the previous one, but I don't remember the precise characters of all messages), and you've asked ngn about ffi plenty of times before

, to the point that I've gotten tired of discussions about ffi

@dzaima (that wasn't a "without the previous message it'd be fine", but "that's extra annoying")

@nathanrogers anyways, back to actual on-topic things - games usually require at least some scalar code (if you don't go completely crazy and vectorize everything), and arraylangs can't compete with C/C++ for scalar code