Newest at the top
2025-09-01 01:41:01 +0200 | <monochrom> | That whole "function binding vs variable binding" business. Perl repels me for the same reason. |
2025-09-01 01:40:25 +0200 | <EvanR> | or the frankenlanguage purescript xD |
2025-09-01 01:39:58 +0200 | <monochrom> | Not bothering. Lisp has other things that repel me. I will just use Haskell or Javascript. |
2025-09-01 01:39:18 +0200 | <twb> | CL is "static typing later" not "static typing never" |
2025-09-01 01:39:02 +0200 | <twb> | monochrom: you can declaim types in lisp |
2025-09-01 01:38:55 +0200 | merijn | (~merijn@host-vr.cgnat-g.v4.dfn.nl) merijn |
2025-09-01 01:38:10 +0200 | <monochrom> | I feel like Python is not too badly write-only. |
2025-09-01 01:37:46 +0200 | <monochrom> | Perl |
2025-09-01 01:37:39 +0200 | <EvanR> | ruby |
2025-09-01 01:37:37 +0200 | <EvanR> | PHP, python, etc, is write only code |
2025-09-01 01:37:24 +0200 | <jreicher> | You could even do nondeterministic code that way. :p |
2025-09-01 01:37:11 +0200 | <monochrom> | s/form/former/ |
2025-09-01 01:37:02 +0200 | <monochrom> | On the upside, since the form just has to be in my head, it can be any arbitrarily undecidable dependent typing. |
2025-09-01 01:36:20 +0200 | <monochrom> | I write typed code in my head, then perform type erasure when I spit out shell scripts or Lisp or Prolog or... |
2025-09-01 01:36:17 +0200 | sprotte24 | (~sprotte24@p200300d16f15cf0019f42a7dc19cb9cd.dip0.t-ipconnect.de) (Quit: Leaving) |
2025-09-01 01:36:06 +0200 | <jreicher> | Writing something that doesn't work is a valuable learning experience. |
2025-09-01 01:35:52 +0200 | <twb> | and also anything over 80 bytes long must be strongly typed |
2025-09-01 01:35:47 +0200 | <jreicher> | And that's pretty much why it's OK for a first programming language. |
2025-09-01 01:35:17 +0200 | <twb> | yeah exactly |
2025-09-01 01:35:10 +0200 | <jreicher> | twb: My rules is weakly typed code is OK if it's OK for it to not work. :) |
2025-09-01 01:35:04 +0200 | <twb> | But I wouldn't write scripts in it that were good enough to commit in git |
2025-09-01 01:34:55 +0200 | trickard_ | (~trickard@cpe-53-98-47-163.wireline.com.au) |
2025-09-01 01:34:50 +0200 | <twb> | so e.g. I do not replace bash with ghci for interactive login shell |
2025-09-01 01:34:42 +0200 | trickard_ | (~trickard@cpe-53-98-47-163.wireline.com.au) (Read error: Connection reset by peer) |
2025-09-01 01:34:38 +0200 | <twb> | jreicher: my general rule is that weakly typed code is OK if it's only going to run once |
2025-09-01 01:33:56 +0200 | <jreicher> | twb: I feel like it's a pretty thin line between untyped and types as weak as bash or PHP. |
2025-09-01 01:33:33 +0200 | <twb> | oh you mean "prompt engineering" as in LLM grifts, not "engineers who work quickly and efficiently" |
2025-09-01 01:32:53 +0200 | <EvanR> | people using free LLMs dwarfs programmers |
2025-09-01 01:32:25 +0200 | <EvanR> | I feel like we're already there |
2025-09-01 01:32:12 +0200 | <monochrom> | Soon it will no longer matter. Prompt engineering will be the first and only language for most people, like it or not. |
2025-09-01 01:32:09 +0200 | <twb> | jreicher: "not typed" meaning weakly typed like bash and php, or *literally* untyped, so true, 1, "" and identity: x -> x are all literally identical |
2025-09-01 01:32:00 +0200 | <monochrom> | Sure. Let me fix my sentence then! |
2025-09-01 01:31:28 +0200 | <EvanR> | "my first and only language" vs "the only first and only language" |
2025-09-01 01:31:16 +0200 | <jreicher> | EvanR: FWIW I only hold this opinion for a first programming language. Later languages (and any serious work) should be typed. |
2025-09-01 01:31:06 +0200 | <monochrom> | Oh, Python not boostrapping itself in practice hasn't stopped a whole lot of people form using it as their first and only language. |
2025-09-01 01:30:17 +0200 | <twb> | https://hackage.haskell.org/package/optics-0.1/docs/Optics.html#g:2 loosk relevant |
2025-09-01 01:30:08 +0200 | <EvanR> | as soon as the AI code behind the scenes is written in "prompt" |
2025-09-01 01:29:37 +0200 | <EvanR> | lol |
2025-09-01 01:29:33 +0200 | <monochrom> | Soon it will no longer matter. Prompt engineering will be the first and only language, like it or not. |
2025-09-01 01:29:29 +0200 | <twb> | Dumb question: when https://hackage.haskell.org/package/base64 talks about "optics" and "lenses", what does it mean? I've only seen "lens" in the context of augeas (an ocaml-y thing) configuration file parsers. |
2025-09-01 01:29:17 +0200 | <EvanR> | I wish I had types in retrospect, in many situations |
2025-09-01 01:28:55 +0200 | <EvanR> | disagree |
2025-09-01 01:28:42 +0200 | twb | (~twb@user/twb) twb |
2025-09-01 01:27:59 +0200 | merijn | (~merijn@host-vr.cgnat-g.v4.dfn.nl) (Ping timeout: 260 seconds) |
2025-09-01 01:25:16 +0200 | <jreicher> | sajenim: No, I think Haskell is quite bad as a first language. I don't think anything with such an expressive type system should be a first language. |
2025-09-01 01:24:29 +0200 | ftzm | (~ftzm@085080231172.dynamic.telenor.dk) ftzm |
2025-09-01 01:24:05 +0200 | ftzm | (~ftzm@085080231172.dynamic.telenor.dk) (Server closed connection) |
2025-09-01 01:23:09 +0200 | merijn | (~merijn@host-vr.cgnat-g.v4.dfn.nl) merijn |
2025-09-01 01:21:23 +0200 | segfaultfizzbuzz | (~segfaultf@23-93-74-222.fiber.dynamic.sonic.net) |
2025-09-01 01:19:48 +0200 | ljdarj1 | ljdarj |