'(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

@joseph_h_garvin Maybe laziness and stream fusion? Ceci n'est pas une liste, etc.

Haskell has more list literals per 100k bytes of source text than tuple literals, while Ocaml has more tuple literals than list literals. Why?

@LukasHozda apprise autolith about its own ccl so it can modify that too in situ 👀

The moment when Autolith finished upgrading how it renders its tools and just started using it immediately in the same session

Even when the LLM is writing Coalton, it still drops down to Common Lisp as needed. It just doesn't need a lot of dynamic typing ceremony when it does it, I guess because it has sorted out that that's the job of the Coalton type system.

With Common Lisp it tends to write a lot more slop code, especially overzealous data predicates.

For nontrivial programs, it seems LLMs converge to working solutions faster with Coalton than Common Lisp, despite there being almost no stable Coalton code in training. Only reason I can discern is that the type system steers it better, and it can "think" as a {Haskell,OCaml}er.

@J7UiBMjH8b I'm not saying b2b saas APIs are fun to True Programmers (TM), but there are definitely lots of programmers who really don't find programming/computers/etc. interesting beyond the surface.

@stylewarning It is mostly a skill issue. She's top talent and can work on whatever interesting idea she happens to have. Of course programming is fun. Yet, the majority of programmers need to churn out the same old b2b saas apis

I don't mean to be flippant when I say this but, yes, people like maximizing money whilst minimizing effort. Other economic factors aside, if software engineers were paid like pure mathematicians, trust me, the people in it would only be in it for the love of the game.

@usgraphics @youwillmakemaps lol @ the middle english spellings

@youwillmakemaps Everytime I see Ordnance spelled correctly, it always feels wrong lol

The 70s......

@yawaramin @eatonphil It's a very very noisy correlation at best. Some phenomenal open source code on a shoestring budget, and some colossally awful code promulgated by trillion dollar companies.

@eatonphil You don’t see any connection between not having any business/monetary/customer constraints and code quality? 😅

I mostly see the Bun and Zig situation the same way Ray described, if you'd actually like to understand it. I would differ in not mentioning TigerBeetle as a Zig flagship because their standards are so strict, and so few teams could do this, and basically no other Zig teams do (including the Zig team themselves), that the fact that they use Zig basically doesn't matter. Ghostty is probably a better flagship to talk about if one must.

@almighty_lisp don't forget about the GOAT of keyword arguments :JUNK-ALLOWED

:FROM-END MAN WHAT A KEYWORD

it's especially bad in technical subreddits. funny enough, some of the artistic ones still show a shred of antislop.

reddit further sloppifies as people post and answer to one another with AI. "They're my ideas, AI just helped me with the English. And honestly? You were right to push back on that. But this point is load-bearing. It's not about the author, it's the message."

LLMs love to prepare you for how honest they're about to be.

@joaompfe Coalton has HKTs and kind inference. (define-class (MyFunctor :F) (map1 ((:a -> :b) -> :F :a -> :F :b)))

So Haskell and Scala are the only languages with higher-kinded types? Cannot be.

@DanielW_Kiwi coalton

Is there a functional language that has a good type system and fast compile times?

@GrowlerEnjooyer @SheriefFYI (I don't like Zig, but there's a bunch of people who are purposefully mischaracterizing what was written to fit their own narrative or their own emotional response to it.)

@GrowlerEnjooyer @SheriefFYI I think the actual argument is more like: If the use of Zig was itself hasty and slapdash, then one should look inward just as much as one looks at their tools for why something didn't work as well as expected. Instead it's painted solely as a deficiency of Zig.

Andrew Kelley changed the content of his blog post, which adds a whole new layer of sleazy. old vs new:

@denbvk lol

@stylewarning I am on the right part of a spectrum which is “Go is actually good”. Haskell and Lisp are the peak of this language enthusiast larping.

Common Lisp is so niche that trillion dollar companies aren't even talking smack about it. :(

@denbvk you don't sound very enthused by languages

@stylewarning As language enthusiast I will tell you: lisp sucks. Its whole niche is meta programming and it’s overrated. Outside rare specific cases like framework DSLs it must not be used. Project uses macroses widely => languages in a language => congrats, it’s unmaintainable slop.