IRC Log - 2025-04-01 - ##forth

Channel: ##forth
Total messages: 45
Time range: 00:34:37 - 23:59:26
Most active: cleobuline (19), mforth (7), veltas (7)
00:34:37 ##forth <cleobuline> mforth: " une fée dans un paysage enchanté " IMAGE
00:34:53 ##forth <mforth> https://tinyurl.com/22aadfol
00:35:55 ##forth <cleobuline> a touch of modernism in an old school program :)
00:53:38 ##forth <cleobuline> mforth: " un groupe de zombies dans un paysage post apocalyptique " IMAGE
00:53:52 ##forth <mforth> https://tinyurl.com/28wepnyo
04:40:07 ##forth <cleobuline> mforth: " un groupe de zombies dans un paysage post apocalyptique " IMAGE
04:40:19 ##forth <mforth> https://tinyurl.com/29usqjhu
05:33:32 ##forth <nmz> what's the fastest forth?
05:34:05 ##forth <nmz> if I wanna number crunch, you'd think something written in assembly would beat a JIT
05:34:28 ##forth <cleobuline> yes
05:38:09 ##forth <cleobuline> forth: " une fée dans un paysage enchanté " IMAGE
05:38:56 ##forth <cleobuline> mforth: " une fée dans un paysage enchanté " IMAGE
05:39:11 ##forth <mforth> https://tinyurl.com/2xttv9ef
05:41:32 ##forth <nmz> cleobuline: like?
05:41:47 ##forth <cleobuline> like what ?
05:42:01 ##forth <nmz> you said yes to my question or no?
05:44:23 ##forth <cleobuline> do you plan to write an assebly forth ?
05:50:52 ##forth <nmz> no
05:51:12 ##forth <nmz> I was whining about forth being slower than C and even JIT languages and someone mentioned vfx forth being the fastest forth, but since I can't even download it, I was wondering what everyone thinks
08:27:16 ##forth <veltas> nmz: If you want to number crunch, then the fastest language is usually literally anything that lets you write SIMD instructions or intrinsics, and spawn multiple threads
08:27:43 ##forth <veltas> If it's really not parallelisable then usually it's still whatever lets you write inline assembly
08:27:52 ##forth <veltas> So assembler, C, Rust, Forth, etc
08:28:35 ##forth <veltas> But if that's not feasible for most of the work, then you're probably better off with C, which has a more mathsy syntax anyway and a lot more compiler effort behind it
08:29:05 ##forth <veltas> Forth is that tradeoff where having a smaller implementation is desirable, but obviously that means you won't get the fastest code in all situations
13:52:36 ##forth <cleobuline> mforth: " un fée dans un paysage enchanté " IMAGE
13:52:49 ##forth <mforth> https://tinyurl.com/2b6xf95h
13:53:30 ##forth <xentrac> it went a bit overboard with the fairy
13:54:19 ##forth <cleobuline> the magic forth :)
13:54:25 ##forth <xentrac> I count onze fées but I might be missing some
13:55:18 ##forth <cleobuline> i like fairys
13:57:06 ##forth <cleobuline> xentrac only 3 fairys
13:57:44 ##forth <cleobuline> mforth: " un monstre horrible avec des grandes dents des gros yeux et plein de poils " IMAGE
13:57:53 ##forth <mforth> https://tinyurl.com/22y5e93t
13:58:17 ##forth <cleobuline> :)
13:58:53 ##forth <cleobuline> don't take lsd with this bot , use at your own risks !
14:21:41 ##forth <xentrac> I think I should use this image as a selfie on dating sites
15:27:26 ##forth <cleobuline> haha
18:20:31 ##forth <cleobuline> mforth: WORDS
18:20:31 ##forth <mforth> USERNAME .S . + - * / MOD DUP DROP SWAP OVER ROT >R R> R@ = < > AND OR NOT XOR & | ^ ~ << >> CR EMIT VARIABLE @ ! +! DO LOOP I WORDS LOAD CREATE ALLOT ." CLOCK BEGIN WHILE REPEAT AGAIN SQRT UNLOOP +LOOP PICK CLEAR-STACK PRINT NUM-TO-BIN PRIME? FORGET STRING " 2DROP IMAGE TEMP-IMAGE
18:20:53 ##forth <cleobuline> new : IMAGE and TEMP-IMAGE
22:45:59 ##forth <pgimeno> I was thinking that with a stack of arbitrary precision numbers, you don't need a string stack - you just need to interpret numbers as strings
23:05:20 ##forth <veltas> Yeah you don't even need a stack, you can just encode everything in one number
23:05:47 ##forth <veltas> But not much point
23:35:59 ##forth <veqq> How would you implement a concatenative language in Forth, just conceptually
23:59:26 ##forth <crc> veqq: is forth not already concatenative?