2024/10/07

Newest at the top

2024-10-07 03:20:21 +0200merijn(~merijn@204-220-045-062.dynamic.caiway.nl) merijn
2024-10-07 03:15:56 +0200ethantwardy(user@user/ethantwardy) ethantwardy
2024-10-07 03:09:20 +0200merijn(~merijn@204-220-045-062.dynamic.caiway.nl) (Ping timeout: 252 seconds)
2024-10-07 03:07:13 +0200 <geekosaur> the ones that aren;t that are mostly cmm code
2024-10-07 03:07:02 +0200 <geekosaur> (the answer is: `+#` and the backend emits the actual instruction)
2024-10-07 03:06:45 +0200spew(~spew@201.141.99.170) spew
2024-10-07 03:06:37 +0200 <geekosaur> how exactly do you write the platform's add instruction in haskell code?
2024-10-07 03:04:57 +0200weary-traveler(~user@user/user363627) (Remote host closed the connection)
2024-10-07 03:04:49 +0200merijn(~merijn@204-220-045-062.dynamic.caiway.nl) merijn
2024-10-07 03:03:16 +0200 <yin> hits the magic way too soon
2024-10-07 03:02:11 +0200fmira(~user@user/fmira) (Ping timeout: 260 seconds)
2024-10-07 03:02:10 +0200 <yin> i wish all code was haskell all the way down to binary
2024-10-07 03:01:10 +0200 <yin> ok great, thanks
2024-10-07 03:01:09 +0200 <geekosaur> necessarily
2024-10-07 03:00:58 +0200 <geekosaur> the actual definitions are baked into the compiler or backend
2024-10-07 03:00:46 +0200 <geekosaur> everything in ghc-prim is "defined as itself", yes. it's just for documentation
2024-10-07 02:59:33 +0200ethantwardy(user@user/ethantwardy) (Ping timeout: 248 seconds)
2024-10-07 02:59:24 +0200 <yin> is it just "magic" and we need the definition for the type signature?
2024-10-07 02:58:50 +0200 <yin> what does it mean for a function# to be defined as itself?
2024-10-07 02:58:09 +0200 <yin> what am i looking at here? https://hackage.haskell.org/package/ghc-prim-0.11.0/docs/src/GHC.Prim.html#readTVar%23
2024-10-07 02:53:56 +0200merijn(~merijn@204-220-045-062.dynamic.caiway.nl) (Ping timeout: 252 seconds)
2024-10-07 02:53:32 +0200EvanR_EvanR
2024-10-07 02:49:24 +0200merijn(~merijn@204-220-045-062.dynamic.caiway.nl) merijn
2024-10-07 02:47:04 +0200xff0x(~xff0x@2405:6580:b080:900:16eb:2432:285b:7ea6) (Ping timeout: 260 seconds)
2024-10-07 02:46:44 +0200Smiles(uid551636@id-551636.lymington.irccloud.com) Smiles
2024-10-07 02:38:32 +0200merijn(~merijn@204-220-045-062.dynamic.caiway.nl) (Ping timeout: 252 seconds)
2024-10-07 02:36:21 +0200Inst(~Inst@user/Inst) (Ping timeout: 246 seconds)
2024-10-07 02:33:58 +0200merijn(~merijn@204-220-045-062.dynamic.caiway.nl) merijn
2024-10-07 02:32:40 +0200peterbecich(~Thunderbi@syn-047-229-123-186.res.spectrum.com) (Remote host closed the connection)
2024-10-07 02:31:43 +0200Inst(~Inst@user/Inst) Inst
2024-10-07 02:30:44 +0200Inst(~Inst@user/Inst) (Ping timeout: 255 seconds)
2024-10-07 02:23:22 +0200fmira(~user@user/fmira) fmira
2024-10-07 02:23:03 +0200merijn(~merijn@204-220-045-062.dynamic.caiway.nl) (Ping timeout: 252 seconds)
2024-10-07 02:22:57 +0200fmira(~user@user/fmira) (Remote host closed the connection)
2024-10-07 02:22:42 +0200fmira(~user@user/fmira) fmira
2024-10-07 02:21:41 +0200supercode(~supercode@user/supercode) (Quit: Client closed)
2024-10-07 02:18:33 +0200merijn(~merijn@204-220-045-062.dynamic.caiway.nl) merijn
2024-10-07 02:18:00 +0200ystael(~ystael@user/ystael) (Ping timeout: 272 seconds)
2024-10-07 02:14:50 +0200machinedgod(~machinedg@d50-99-47-73.abhsia.telus.net) machinedgod
2024-10-07 02:07:59 +0200merijn(~merijn@204-220-045-062.dynamic.caiway.nl) (Ping timeout: 260 seconds)
2024-10-07 02:03:03 +0200merijn(~merijn@204-220-045-062.dynamic.caiway.nl) merijn
2024-10-07 01:57:51 +0200myxos(~myxos@syn-065-028-251-121.res.spectrum.com) myxokephale
2024-10-07 01:51:57 +0200merijn(~merijn@204-220-045-062.dynamic.caiway.nl) (Ping timeout: 252 seconds)
2024-10-07 01:49:17 +0200 <L29Ah> implying llms can utilize boring things like file format specs
2024-10-07 01:48:38 +0200 <L29Ah> true programming laziness
2024-10-07 01:48:12 +0200 <L29Ah> so ideally one converts a PC into a heater that hopefully produces working libraries once in a while :>
2024-10-07 01:48:11 +0200 <Inst> but afaik Haskell generated by Agda is unreadable
2024-10-07 01:48:03 +0200 <Inst> i'm wondering if LLM would have an easier time converting a domain into Agda types, then using Agda facilities to pound out a Haskell program
2024-10-07 01:47:37 +0200merijn(~merijn@204-220-045-062.dynamic.caiway.nl) merijn
2024-10-07 01:47:27 +0200 <L29Ah> and one can pester llm to produce test cases too to catch all the stupid errors programmatically