File: 68794ca.png (274 KB, 786x1080)

274 KB PNG

>>110144513 >>110144643 >>110145747

Don't know why these threads are so short-lived all of a sudden, but let's try again.

File: [email protected] (2.56 MB, 2256x1504)

2.56 MB PNG

guix

>>110144306
What use is a good programming language when nobody even programs anymore

>>110144306
I'm a Clojure fag and too busy writing code. most people discuss Emacs here which idgaf about (I use (Neo)Vim btw). I was planning to comment on the stupidity of static types in (most) situations in the 2nd-to-last thread, but I figured it is a waste of time, because the same shit will be discussed in a week.
Lisp is good. Common Lisp has great ideas regarding interactivity, and really BTFO everything else for a couple of decades. I think Clojure is better now. I think an even better lang could be devised. Idk if sexps are the ultimate syntax, PURELY because of accessibility concerns regarding indentation and the discernment of a literal list vs. a function call is not possible in a macro without context (this also makes code formatting difficult).
idk what else to say...
Lisp is great and most people here asking about it won't even spend an hour trying to solve an AoC problem, so what's the purpose?

>>110144301
based Untitled 3 enjoyer

t. lover through peak hipster times in 2010

>>110144301
Grandpa ?
Is that (you) ?

>>110144513
Well, if you don't use your brain, you lose it. Just because some people seem to be okay with that these days doesn't mean that you should be.

>>110144643
>I was planning to comment on the stupidity of static types in (most) situations
What's the hot take? Never heard an argument *for* dynamic types, only for static ones.

emac is dead and we have killed it

>>110139551
I use tramp very sparingly due to its tendency to block the Emacs UI. When I do use tramp, it's usually through eshell. I think it's because I find the blocking more tolerable in the context of an interactive shell. It's less upsetting to me somehow.

>>110145250
nta but there are better ways to take care of your brain than coding, you may as well be doing something useful.

anyway on the side i like to use llms to generate anki decks.

>>110144306
You have to format the subject a certain way for them not to slide off the board. Look at how all the other generals are formatted, you think that's just an unspoken standard? We can't even coordinate enough to spell a word. Watch

N

>>110145747
I

File: GWjEGEoXgAYDtxT.jpg (299 KB, 900x1200)

299 KB JPG

>>110146959

>>110145780
G

>>110145599
Dynamic vs. static types has been an argument that has gone around in circles for decades, so this tells me you're just young.
If you want to spend some time on old arguments, peruse https://wiki.c2.com/?BenefitsOfDynamicTyping and further links from there.
Empirically there's no strong difference at scale, plus there are more dimensions to a type system than simple static or dynamic, plus there are more dimensions to a language than its type system. Some of these extra properties will likely matter a lot more for whatever it is you're trying to build.

>>110146083
G

>>110146959
A

File: 1780714780482422.png (6 KB, 710x94)

6 KB PNG

>>110153933

>>110144301
why does emacs suddenly show search preview when I do C-s?
how do I disable this? normal behavior makes the cursor jumps around the file buffer

>>110146164
>Dynamic vs. static types has been an argument that has gone
It looks like a done debate today because of vibeshitting.
LLMs make random small mistakes or change the code at whim (even when you ask it not to) so strong static types seem like a most
See how Rust and Typescript are the most popular languages on GitHub despite being considered pain to write before 2023

>>110147474
*Must

>>110147474
The argument was over before LLMs. See how every modern language has a static type system and older languages that didn't have been retrofitted to support them (python type hinting, ruby sorbet, typescript).
The obvious reason is that it's hard to know what a variable is a lot of the time in a dynamic type system, and with type inference or gradual typing and modern LSPs, having types is more ergonomic than not.
LLMs didn't fundamentally change anything here, they just amplified the existing relationship.

>>110144643
>I was planning to comment on the stupidity of static types in (most) situations in the 2nd-to-last thread, but I figured it is a waste of time, because the same shit will be discussed in a week.
Please do. I'm curious for arguments for dynamic types because I have never heard an argument that was convincing enough. It's always something orthogonal to static vs dynamic typing *in principle*.

anyone have any opinions on https://autolith.rocks/ or other lisp based coding agents?

>>110144513
I was about to ask the same kek. There is literally no purpose. In a couple of years AI will write basically perfect code, and there will be 0 reasons to even review the code (we are kinda there right now to be honest). All the rest is cope.
Furthermore, there is no point in text editors too, as AI does all the editing. So discussions about emacs are worthless too.

>>110148372
text editors are for things other than just writing code. Unless you think that people are going to promp in a little box piece by piece forever like they were using ed

>>110145686
>nta but there are better ways to take care of your brain than coding, you may as well be doing something useful.
Such as? I think you can take care of your brain in other ways than coding, but all those ways too are useless. For example math.
Also, what to do while working? Using ai for 8 hours a day will inevitably turn your brain to shit.

>>110148384
AI can manage its own prompts too. It can write software end to end without human intervention.
Other than prose literally what's the use case of something like emacs or vim? Even auxiliary tasks like managing git repos with magit could be done by ai. Everything that can be done in a computer could be done by ai, especially code related or sysadmin related.
The only reason to use a text editor is to manage notes, write prose / poetry or manage one's own agenda.

>>110145670
>I think it's because I find the blocking more tolerable in the context of an interactive shell. It's less upsetting to me somehow.
Yes I found this too. It also tends to block less often because there are less buffers involved i think.
The problem is that dired with tramp is just too comfy. And I sometimes need to edit files remotely, so even if starting from eshell, I'd do find-file my-remote-file.txt.

>>110148412
so true, I cant wait for the full-brain neuro-link so that the machine can just preempt what I want built without me having to write anything. Or maybe we can just write unstructured text and argue with llms line by line and read paragraphs of claude-slop until we have conveyed our ideas enough to be implemented.

Forth > lisp
/thread
inb4 op can't /thread
inb4 op can't inb4

>>110147515
>pajeet likes it so it must be correct
Poo in loo, not in thread

>>110144643
Can you share your neovim setup for clojure? Not the actual dots but a high level overview.
I'm new to both clojure and neovim. Right now I've set up clojure-lsp with neovim and I'm using conjure (+ treesitter + vim-dispatch + vim-jack-in) to evaluate stuff, but I've not set up good s-expr edditing.

>>110147515
python type hinting is not a type system, it's glorified comments. There's a big difference when developing your application, because if you want to change the functiion argument from, say, dict[str, dict[str, int]] to something else you can just change the function body + how it's called (even if you already have type hints in place and the LSP is throwing warnings).
Excessively type hinting your python code at first pass (when you're not done / ready for some release) is peak brainlet behavior, similar to writing docstrings at the same time.You're spending time not writing code AND writing comments than may be obsolete in a few days or even tomorrow.

>>110149023
I don't like python hints, but I also like them.
Fastapi makes heavy use of them, I don't agree with all things there but they are nice.

Maybe it is because I am maintaining a legacy python 2 piece of shit and just now we are migrating.

>>110149528
Python 2 is objectively superior to the mess that is python 3, and you will soon understand it.

>>110149634
this is the most retarded thing I've ever read. Did they close down /dpt/? usually the languagephiles stayed there.

>>110148412
>The only reason to use a text editor is to write prose / poetry
but ai can write poetry too, anon

>>110150239
Yes it can, but you usually do these things to express yourself. Code is purely utilitarian.

>>110150249
>you usually do these things to express yourself
what's the point in that

File: 10058083.png (313 KB, 1230x896)

313 KB PNG

>>110150239
>ai can write poetry too

>>110149732
It doesn't even have to do with language design. But it's OK timmy, some day you'll have done more than fizzbuzz and you'll understand why needing a team of 10 dedicated to upgrading your entire system for each sub-point version of python 3 is not great.

File: cython.png (40 KB, 960x429)

40 KB PNG

>>110150579

>>110150530
Hopefully Python 4 will run on Cython

>>110150561
Hopefully not. Cython is another abomination. It's just c-flavored python.
Hopefully, python 4 is not a piece of shit, but there's too much fundamentally wrong about python to really entertain the hope, especially since 3 didn't learn at all from the 2-to-3 transition nightmare, is slower than 2.7, and overall, shows that guido and palls don't like learning from their mistakes.

>>110150579
Python 2.7 is my second favorite language after Common Lisp, only narrowly beating out Clojure. Any version of Python 3 is like not even in my top 10. I doubt Python 4 will improve things.

>>110150649
I don't see how. Python 2.7 was also a shitlang, it just remains the only good pyshit version today is all. Otherwise, it is slow, is stuck with significant whitespace, has amateur hour-tier implementation of common data structures, notably dicts, and the ecosystem is composed largely of researchware and studentware with few quality software that would nevertheless benefit greatly from a skilled engineer cleaning them up.

It's a very good replacement for bash scripts, but not for bash itself (significant whitespace on the command line is cancer), and isn't a particularly good glue language compared to, for example, lua. It isn't particularly fast to develop in like lisp is, nor fast to run in spite of fast development.

I'm no fan of clojure either, but unlike common lisp which has an anemic ecosystem, the python advantage there gets gap-closed in clojure, while retaining several useful properties, if castrated, of these other example comparison languages.

Objectively, clojure is superior to any version of python.

>>110150703
The dict implementation was fine, elegant even, compared to many other hash table designs of the era. The only trouble was perhaps overuse of them, but they were a joy to use.
Speed is a property of an implementation, not a language. It was fast enough for many things, and could easily call out to foreign code when not. And not just C code, but C++ code too. We Python devs had access to all the C++ goodies as much as C goodies. (And if we wanted Java goodies, there was Jython.)
Significant white space is a good thing in a language. Not so much in declarative markup, yaml can die.
Clojure is very good but also slow, and too opinionated with its emphasis on immutability and laziness. It's much better on concurrency matters, which was another big Python weakness, though again we made due. (Twisted was the goat.)
https://www.norvig.com/python-lisp.html offers some other perspectives.

>>110150783
What makes you think Clojure is slow?

>>110148965
I highly recommend Parinfer. VERY comfy.
https://github.com/eraserhd/parinfer-rust
I also recommend
https://github.com/tpope/vim-sleuth
with a good .editorconfig so that useless whitespace is deleted on buffer write to file.

>>110150783
>The dict implementation was fine, elegant even
Come on, this is far too obvious a bait. Not even trying-tier.

>>110150830
Experience. It's relative slowness, of course, I don't think there's any situation today I'd use Python where Clojure wouldn't work from a performance standpoint, and probably the majority of those Clojure would perform better. Maybe I'm still just sore about all the times we had to optimize Clojure by tearing out the nice persistent data structures and replacing them with ugly Java interop code, or even worse rewriting large sections in Java directly. When you do enough of that, you start wondering why you're even bothering not just keeping everything in Java, which on its own terms is not a bad language anymore.
>>110150869
https://github.com/python/cpython/blob/2.7/Objects/dictobject.c
https://github.com/python/cpython/blob/2.7/Objects/dictnotes.txt
Or see Chapter 18 from the book Beautiful Code, "Python's Dictionary Implementation: Being All Things to All People".

>>110150937
Bad bait is horrid

>>110150937
I'm not sure what you're on about, Clojure can be as fast as Java, as of https://github.com/joinr/performancepaper
It's true it may not be the most idiomatic code but the point is you have the choice and gradient of how you want your code to look like.
But if it's SLOW you're doing something terribly wrong, like using `merge` or `satisfies?`
Did a ton of performance work on Clojure, some internally and some shared with the wider public.

File: depressed wizard.gif (1.26 MB, 360x270)

1.26 MB GIF

>>110151919

>>110150984
NTA, but at least that anon tried to contribute something to the conversation.

>>110151837
There are no points for trying, and even fewer points for making shit up. Points are only accorded for results, facts, or developed, grounded opinions.

>>110151919
>Points are only accorded for
This is not Reddit. If you want to upvote or downvote posts based on whether you feel like you agree with them, you can go somewhere else.

>>110152001
This is indeed not redd!t. Thus, we do not give points for posting slop, human or ai.
>>>/r/eddit/

File: 1784172152823545.png (1.06 MB, 1489x1421)

1.06 MB PNG

>>110152115 >>110155484 >>110157141

>>110147474
>See how Rust and Typescript are the most popular languages on GitHub despite being considered pain to write before 2023
Source? I see Javascript and Python.
>so strong static types seem like a must
Source? The paper I saw didn't show this.
They're investing in formal verification:
https://www.anthropic.com/research/formal-verification
Strong static types just seems like placebo.

>>110152092
>Strong static types just seems like placebo.
They are.
Virtually every piece of research about them shows this to be true.

File: carmack.jpg (166 KB, 1920x1080)

166 KB JPG

>>110152338

>>110150579
>3 is slower than 2.7
Well, what if we drastically improve the performance by making it possible to compile Python to another language, such as—
>Cython is just a c-flavored abomination.
Well, never mind, then. There go my hopes for efficient loops, I guess.

Pic related (around 13:00):
https://www.youtube.com/watch?v=I845O57ZSy4

>>110151473
For Clojure, it "can be" as fast as whatever, but the costs are often high to get there. Greater expertise, less idiomatic code, harder to debug code, and worst is, like Haskell, there are cases where the result is fragile and someone coming in later writing otherwise normal Clojure code can inadvertently ruin the whole effort. Transducers hurt more than help. Decisions have non-local effects.
You'll have a better time writing things in Java, or even JavaScript, or Common Lisp if you're lucky, and you'll keep more idiomatic code that's more robust to future changes.
I know Clojurists will debate at least some of that, and of course for much ordinary software the performance of Clojure is fine. (The performance of Python, Ruby, Perl, PHP, JS, and many other languages is also fine.) I'll just point to the public record and note that there are years and years of posts from individuals seeking guidance on why their Clojure code wasn't performing as well as they hoped, large systems like Apache Storm being rewritten to Java, many reports of medium to large applications becoming Node.js applications, and people dropping Datomic. There's a long reputation of slowness that isn't without some merit or entirely confined to start-up times or short lived programs.

>>110152274
I should correct myself, I meant it is a python-flavored c abomination, rather than a c-flavored python. The problem is you end up writing a c program but with significant indent, losing both advantages. Plus the toolchain is dogshit, packages using cython released without binaries fail to build 80%+ of the time.

For speed I try to pack things and use numpy (e.g. numpy structured array when relevant and copious use of ufuncs), or numba which is kinda a mixed bag, but if you can satisfy it it can really make some things fly. Total crutches, though.

Python also has 3 (4?) different ffi systems, all of which are broken in different ways. Some literally don't work, some only support a specific type of symbol, and some have such stupidly high overhead that you get faster code by just writing python than writing a small C function that includes your hot loop. As a result, the only reasonable way to get speed in python is to write actual C code (so you don't have to deal with the cython toolchain and disgusting syntax) using python.h yourself, then it can be loaded directly in python without an extra ffi layer. However, this is incredibly error-prone compared to writing normal C code because the python.h interface is total trash.

>>110144301
is Emac the future of software?

>>110152336
Everyone with a brain knows clojure is slow as balls. Still 20x faster than pyshit though.

>>110152336
>>110152361
Clojure is, on average, slower than Java, (which isn't all that slow), and that's usually by less than one order of magnitude, which is NOT BAD
way better than JS, and still waiting for good, modern comparison on Clojure and CL performance.
I'm familiar with some of those stories. The solution is often trivial. It took a while to figure out what idiomatic Clojure code looks like but you can write Clojure code with good performance within a set of idioms still considered Clojure-ish, and when you deviate from them, still very readable. Often some moron used `satisfies?` and ruined everything forever.
What IS true and annoying is that Clojure comes with a ton of performance foot guns - merge, merge-with, reflection, laziness. You can't fix them, you just have to avoid them.

>>110152502
>Clojure is, on average, slower than Java, (which isn't all that slow), and that's usually by less than one order of magnitude, which is NOT BAD
this is just disingenuous, we've had multiple massive prod issues because of misuse/abuse of lazyness leading to enormous slowness and memory bloat, which just straight up wouldn't happen with java because java devs aren't allergic to just writing normal loops
yes, you CAN write performant clojure, in reality many clojure devs simply do not
transducers?
>uuhhh that's so hard to read and complex
loop/recur?
>not idiomatic!
function local transients?
>eww mutable state
and so on
i've given up on trying to introduce these concepts into our projects

>>110152550
We've had similar issues. Some teams were poorly managed and run by filthy animals who couldn't adapt, other teams excelled and everyone was using transducers after I grabbed them by the ear and drilled them on the subject.
The problem you're referring to, I think, is not with Clojure the language / performance / design, etc, but a degenerate social state which arises in Clojure, and I experience it too at $WORKPLACE, which is that most devs just don't care. They're there to do their day job and go home.
I don't know how Nubank solved it but I don't see large Clojure organizations collapsing under their own weight when cultural drift erodes their technical foundations.
I had about a 2 hour chat on the subject with Clojure's dev advocate. Can't tell you what I expected to come out of it, but least I could do is give him my time, I still want the language to succeed.

>>110152736
>I don't know how Nubank solved it but I don't see large Clojure organizations collapsing under their own weight when cultural drift erodes their technical foundations.
Wagie in the cagie. e.g. Nuslavery.

>>110152502
tf do you mean you shouldn't use merge?

>>110147474
>static types are good for (bio)llms that can't think
We know.

>>110152115
Source?

>>110153861
NTA but it's well known that any time the subject is studied the results are inconclusive or very weakly in favor of one or the other for some context. If it mattered, the results would be stronger even with study design issues.
https://danluu.com/empirical-pl/

>>110147046
it's probably from a package you installed. i don't think vanilla emacs has that functionality

i switched to the doric themes and i realized that real syntax highlighting actually really fucks with my concentration

>>110152092
Every single fucking language is "the best" for AI.
>C is the best, it's the fastest so it's the best
>Python is the best, big corpus, dynamic so less code which means less errors, it's the best
>Rust is the best, types means statically caught mistakes, it's the best
>X is the best, it does Y
>anyways AI is the best ever and will go to the moon, ask it to spit out assembly directly

>>110148412
>AI can manage its own prompts too. It can write software end to end without human intervention.
>Even auxiliary tasks like managing git repos with magit could be done by ai. Everything that can be done in a computer could be done by ai, especially code related or sysadmin related.
Yeah and AI will be able to fuck your wife for you too

File: roll.png (106 KB, 262x250)

106 KB PNG

>>110155484
Someday neural networks will program your computer through a lengthy fart sound that is compiled to the world's most horrific assembly.

>>110154742
>real syntax highlighting actually really fucks with my concentration
why?

2 emacs questions

1. What is the easiest way/commands to setup a C/C++ development environment? I just have lsp and clang installed and enabled and already it's a pain autoformatting everything and any time I look online there's a rush of shit I'm barraged with that all looks like it conflicts with each other, like electric indent, indent size, c-style etc.

2. You could maybe answer if you use obsidian or another note thing because I think they're similar, how exactly are you supposed to use org roam and why wouldn't I just have one giant file I can search through? Every time I look into org roam its benefits don't really make sense. It mainly looks like you just make these webs of info that look nice but take all this manual linkage and planning which is almost something I would see useful if not for the fact that search functions can find what I want pretty much instantly, like I could just tag stuff with headers and flick to the next header that has a keyword

>>110156135
>What is the easiest way/commands to setup a C/C++ development environment?
Have you tried using an Emacs distro like Spacemacs or Doom Emacs? Spacemacs has a c-c++ layer that sets up C and C++ for you.

If you're not into all that, there might nevertheless be a few clues to be gleaned from the documentation:
https://www.spacemacs.org/layers/+lang/c-c++/README.html

>>110156135
>electric indent, indent size, c-style etc.
https://www.emacswiki.org/emacs/IndentingC

>>110153933
no, it's emacs builtin feature
I tested with

$ emacs -q

>>110153354
merge is / was very slow (changed in Clojure 13), used to use conj and build up a sequence of maps to merge maps. Was always my first culprit when looking for performance bugs.
now uses reduce over the source with transients. Clearly better but would have preferred a direct kv based path.

>>110153876
Thanks. I also think the problem is that these kinds of studies are invariably a meme. You could "prove" everything and its opposite. It's the same in psychology and other meme fields.

>>110157003
The question you should be asking is who funds these studies in their favour.

>>110152092
https://youtu.be/qnKf_FwqYrU?is=SyMS0KZO9sSukR_m

>>110157141
>This is why Static Typing WON
>doesn't show anything that isn't native to dynamically typed languages also
Yes, this is why static typing won: brainlets who don't know shit about programming and dynamic languages following the pajeet herd.

>>110157141
https://www.youtube.com/watch?v=2J-HIh3kXCQ

>>110158639
Brainlet as always. Blowjob never disappoints.

>>110158678
>J. Blow never disappoints
He uses Emacs and talks about programming, so on that basis alone he is unironically more on topic than some people in this thread.

>>110158639
The lisptranny can't understand this power.

>>110159025
The lisp gods are in their ivory tower looking down on the caveman so proud of himself for discovering how to sharpen a stick using a rock.
The caveman proudly proclaims that his sharp stick will make an unbeatable weapon on the soft and round guns of the lisp gods.

File: photo_2026-10-25_08-48-38.jpg (96 KB, 1280x541)

96 KB JPG

>>110159025
>he thinks that Lispgods have not used other langs before
post code, I bet you think these types assisted me (they didn't, they annoyed me)

File: https:⁄⁄lispcookbook.gith(...).png (97 KB, 860x700)

97 KB PNG

>>110159223 >>110159242

>>110159025
>Lispers can't understand this power
Not this again. Lispers absolutely can see the benefits of this power, which is why some have seen it fit to add it to the language.

File: https:⁄⁄docs.racket-lang.(...).png (88 KB, 780x980)

88 KB PNG

>>110159215

>>110159215
they are pure retards that do not understand that code is data, and they can do whatever they want with the data that is their code. they have no idea of the utility of eval in contexts like these. they think their compiler just MUST exist as a separate binary executable to their program. they will never get it, because they are nocoders or lazy fucks that will never try anything new from what they learned programming was when they were in high school.

File: Syntactic Abstraction.png (394 KB, 2440x1600)

394 KB PNG

>>110150937
>Beautiful Code: Leading Programmers Explain How They Think
Chapter 25 of that book contains a cool essay on hygienic macros by R. Kent Dybvig.

>>110148369
I use hermes.

> autolith
> common lisp
why not guile scheme ffs

File: ranking-total-jobs.png (286 KB, 1554x1862)

286 KB PNG

>>110161529

>>110157141
"won"

>>110160247
>JavaScript / TypeScript
The top contender is a dynamic language to which static typing was added later. (The other way around is harder.)

>>110161529
Indeed, and regular JS has type hint support through JSDoc, Python has optional typing support, PHP has optional typing support, Common Lisp has optional typing support. This suggests that dynamic typing "won", but people do want some of the benefits of static types, sometimes, and it's easy to get them. Even the other way around, people want something closer to dynamic types most of the time.
What language is this?

        var jsonPayload = """
                {
                    "name": "Jane Doe",
                    "email": "[email protected]"
                }
                """;

        var request = HttpRequest.newBuilder()
                .uri(URI.create("https://example.com"))
                .header("Content-Type", "application/json")
                .POST(HttpRequest.BodyPublishers.ofString(jsonPayload))
                .build();

        httpClient.sendAsync(request, HttpResponse.BodyHandlers.ofString())
                .thenApply(HttpResponse::body)
                .thenAccept(response -> {
                    formatResponse(response);
                })
                .join();

(Answer: Java.)

>>110161803
This is Java sir

>>110161899
And yet here's what some old school Java looks like, which is part of why so many of us flocked to dynamically typed languages

    final String jsonPayload = "{\n" +
        "    \"name\": \"Jane Doe\",\n" +
        "    \"email\": \"[email protected]\"\n" +
        "}\n";

    final ExecutorService executor = Executors.newFixedThreadPool(1);
    Future<String> futureResponse = executor.submit(new Callable<String>() {
        @Override
        public String call() throws Exception {
            HttpURLConnection connection = null;
            try {
                URL url = new URL("https://example.com");
                connection = (HttpURLConnection) url.openConnection();
                connection.setRequestMethod("POST");
                connection.setRequestProperty("Content-Type", "application/json");
                connection.setDoOutput(true);

                try (DataOutputStream wr = new DataOutputStream(connection.getOutputStream())) {
                    wr.writeBytes(jsonPayload);
                    wr.flush();
                }

                int responseCode = connection.getResponseCode();
                StringBuilder response = new StringBuilder();
                try (BufferedReader in = new BufferedReader(new InputStreamReader(connection.getInputStream()))) {
                    String inputLine;
                    while ((inputLine = in.readLine()) != null) {
                        response.append(inputLine);
                    }
                }
                return response.toString();
            } finally {
                if (connection != null) {
                    connection.disconnect();
                }
            }
        }
    });

    try {
        String response = futureResponse.get();
        formatResponse(response);
    } catch (Exception e) {
        e.printStackTrace();
    } finally {
        executor.shutdown();
    }

>>110148965
almost forgot, do not forget to tell Neovim that .cljs, .cljc, .bb, etc. are all Clojure files. idr which ones you actually need, I just blind-listed them

>>110162146
No doubt, Java's made significant progress, freely cribbing features from its children, Kotlin, Scala and Clojure

>>110162146
Strictly speaking, it's an expression problem, not a dynamic vs. static types problem. And yet back in the early 2010s and before, almost every static typed language had expression problems making them about as ugly as Java, whereas almost every dynamic typed language had rich expression. As languages have evolved, static type languages have converged close to dynamic languages both in expression and making the type system less noticeable, whereas dynamic languages have added some optional typing features and that's about it. The direction of influence is pretty clear.

>>110162170
No. This is the dividing line, or the point of utility balance, not the crux of the matter itself. The matter at hand is actually productivity, which was treated as RAD vs maintenance. But as you rightfully pointed out, this friction has largely been dealt with and dynamic languages' obvious win has now been gap-closed for the most part. But the main question (productivity) is still not properly solved.

>>110162146
This has nothing to do with dynamic typing vs static typing. Type inference doesn't imply dynamic typing btw. Good morning saar.

>>110148392
>Also, what to do while working?
learn something.
i do my anki deck or play violin whilst the llm is slopping.
>for 8 hours a day
the trick is to do it for just one hour and claim it took 8

>>110162146
>types
This is ENTERPRISE mental illness.
You can write Java in any language.

>>110163141
>i do my anki deck or play violin whilst the llm is slopping.
So nothing useful. Not saying it's bad, but I think there is no useful activity that prevents ai brainrot. They are all useless like math, music or programming.

another ClojureTV talk just dropped
https://www.youtube.com/watch?v=7yeH1DqK984
it's all about AI/LLMs, very interesting/funny so far (I'm only 18 mins in)

Sometimes creating new tab in emacs takes a long time(2 seconds). I didn't tested it much, but it feels like it fixes itself after opening new tabs couple times and will be fine for some time after that.
Anyone here knows why it can be slow?

>>110168950
That's strange. For me, opening a new tab is instantaneous.

Try this:
- M-x toggle-debug-on-quit
- Then try opening a tab.
- If it's taking a long time, hit C-g to try to force Emacs to quit, and then see what kind of stack trace you get back. (Maybe that'll offer a clue on what's taking so long.)

When you don't want to see stack traces after C-g, you can M-x toggle-debug-on-quit again to turn it off.

>>110168950
It's common. Emacs is a buggy piece of shit and the various plugins tend to individually be super buggy and fuck the entire emacs up just by existing.

File: 2026-10-26_09-37-39.png (50 KB, 813x535)

50 KB PNG

>>110169286
Here's a little tip to make Emacs' stack frames look better. It formats the stack frame as sexps so it looks more like the actual elisp code that's executing.

(setopt debugger-stack-frame-as-list t)

>>110144650
>>110144301
not op but yep i like samskeyti too

File: en.png (212 KB, 850x400)

212 KB PNG

>>110175181

>>110172352

>>110166386
>So nothing useful
there is no such things as an useful thing.

>>110166386
>>110173060
rather i should say that there is nothing that is objectively useful, it's all a question of frame of reference.

playing violin or doing my anki deck is useful to me.

1. playing violin makes me better at violin, which is useful for my goals.
2. my anki decks helps me learn more stuff that is also useful to my goals.

File: 1791302879230870.jpg (427 KB, 1860x1376)

427 KB JPG

why the FUCK is "bump" swallowed silently
jannies?!

>>110173050
That used to be true until it became profitable to lie and manipulate information

>>110173080
>>110173060
Ok but at that point you might as well do some programming if you enjoy it. I was replying to >>110145686
>you may as well be doing something useful.

>lisp general is mostly about python and java

I didn't know things could even get this grim for lisp cucks.

>>110175387
Lisp is now vibecoded slop so might as well.
It's too bad, it had some great ideas while it lasted, but they're now doomed forever.

Long live forth!

>>110175442
Forth will always be best served at a temporary REPL, as a human input technique. positional thinking is a temporary convenience that destroys understanding at a later date.

>>110175531
Unfortunately, that is correct. But surely there is a way to build abstractions around a forth core that keeps the advantages and gets rid of the undue mental load there, thus getting us One Language to Rule Them All (OLTRTA).

Emacs ricing can really take up a lot of time, portacle is outdated so i made my own portable emacs common lisp environment which i can easily update myself, just trying to update sbcl breaks portacle

How good are LLMs at Lisp? Has anyone here slopped Lisp in any serious capacity?

>>110175817
If you are going to slop heavily you might as well just use rust. Why bother with Lisp if you won't interact with it anyways

File: ddfc1e00-6638-11e9-9b80-0(...).png (725 KB, 1280x640)

725 KB PNG

>>110175871

>>110175545
>One Language to Rule Them All (OLTRTA).
but julia laready exists

>>110175828
It's just my morbid curiosity, because I found that LLMs are absolute garbage at Lua for example and I wanted to hear from actual Lisp users whether they ran into the same wall.

>>110175817
I tried clojure with glm 5.3 and it struggled with simple programs. Actually, it couldn't even balance parenthesis correctly without special tools. That was the day I learned that lisp is unironically too hard.

>>110175829
Julia is a broken piece of garbage.

>>110175844
Code looks intimidating to me without parens now, like it's just floating there, gross lmao

>>110153876
>If it mattered, the results would be stronger even with study design issues.
That's not how studies work.

Hey OP the reason these threads are probably dying is because it's literally just off-topic bickering about static types. Take the hint and realize these threads aren't welcome here anymore.

>>110175835
I slopped a quick game engine to test sdl2 and sdl3 in common lisp and it worked well

How do I git gud at the repl? Any good books that specifically teach the workflow, how to handle the debugger, things like that?

File: bookslisp.jpg (1.18 MB, 3082x2476)

1.18 MB JPG

>>110177271 >>110212199

>>110176723

>>110176723
Common Lisp Cookbook

>>110177174
Achsully, I'd love to read some GOFAI books. Bonus points for lisp of course. Anything interesting? No toy shit, actually cool stuff. Constraint satisfaction, goal inference, things like that. Not yesteryear's horrible NER. Maybe something with LUSH if there's anything good there.

File: https:⁄⁄atgreen.github.io(...).png (169 KB, 1300x1000)

169 KB PNG

>>110176433
>Hey OP the reason these threads are probably dying is because it's literally just off-topic bickering about static types.
We tend to allow for some mentions of other languages so that we can compare language features and speak sensibly of interlanguage projects like pic related. Otherwise, there's >>>/g/dpt.

>>110177271
I suggest you learn english first

>>110177677
Pajeet, please. Maybe you'll understand better if you cleaned the shit smears off your screen.

>>110178041
Poojoo pubehead cant read lol, keep failing at life

>>110178055
Poo in the loo, not in the thread.

>>110178090
Lol self-hating pubehead cope

>>110148372
There's no purpose in discussing AI because AI will prompt itself. Now fuck off vibejeet.

>>110175337
coding by hand is nowadays generaly less useful than playing an instrument or learning a language, at least if you already know how to code, if you don't there is still value in learning it imo but yea, no point in handwritting most shit anymore.

>>110178371
>coding by hand is nowadays generaly less useful than playing an instrument or learning a language, at least if you already know how to code
Fair point, maybe I will learn some math.

>>110178371
kek
>>110178473
kek

You jeets are so funny

>>110178371
this is true in theory or in corporate-slop, if you want programs that work actually well you need to so it yourself or at least fix the codeslop

>>110178642
>this is true in theory or in corporate-slop
which is the only place i code for anymore, i'm tired of computers, if i have freetime i rather just go touch some grass now.
>if you want programs that work actually well you need to so it yourself or at least fix the codeslop

there is granularity to this, asking the llm "add this struct then that function that takes that as input and output this" is still faster than writting it yourself.

sure i'd not say "add that whole ass feature" but even for stuff you care more about it's still faster to use the llm and just as good as long as you are specific enough.

>>110178701
I can smell the curry and poop from all the way over here. Even the most poojeeted jeet is faster, better and cheaper than the best frontier LLM.

>>110178722
post hands
at no point i said llms were good or better than what i'd write.
they are however much faster, you can't type faster than an llm can output.

>>110178793
Sure I can. They type super slow. Especially when you consider that there is no qualitative difference in the value of the output they provide compared to just hdophfiopasndfoiansdif since they're such trash, though you wouldn't know because you still haven't figured out how to shit in the toilet.

>>110178829
You should spend more time thinking than typing

>>110178829
>They type super slow
no they don't.
you are using the wrong provider.
some can output above 100t/s which is equivalent to 4500wpm, the human world record is about 200wpm
why do you feel the need to lie.

heck even a super slow llm that does 15t/s would still beat human world record.
let alone me or probably you, i type at 120wpm at most.

>>110178845
prolly a brown dude projecting honestly.
he can't even into basic wpm math.

>>110175817
>How good are LLMs at Lisp?
Paid Claude is very good.
>Has anyone here slopped Lisp in any serious capacity?
Yes.

>>110178847
>>110178845
>ignores the 30m """thinking""" time before the first token gets output
Pajeet, please.

>>110150835
I didn't try sleuth but parinfer is amazing.
I programmed Conway's Game of Life for practice and it was pretty fun. Does anyone have a recommendation for another beginner/intermediate project? If I don't come up with anything I'll probably add a GUI and mouse support to my game of life.

>>110178974
1. thinking can be disabled
2. even if enabled you can set it to low.
3. even if set to high, taking into account the latency it'd still be faster than any human.
stop trying to pretend otherwise you just make yourself look retarded.

even if thinking was 95%+ of its output it'd still be faster than humans can type.
and that's assuming you don't use models that can get beyond 100t/s (there are).

>>110168424
A strange talk. Started off strong. I'd like to use this idea of metis more, though merriam-webster's definition of métis made me lol. The vibe acronym was funny but the rest not really. Bashes the long history of crap studies and charts and then tries to use studies and charts to bolster his points. (Ok I guess he mostly beefs with productivity studies, but it's really a larger set of studies that have issues.) Total blindspot on AI's usage at breaking IP monopolies or being useful for self-education. The points about AI etiquette are obvious but unrelated to what I hoped the talk would be about from the title.

>>110179340
>you can make it so retarded that all it comes up with is literally sdlnfsldkfnsd or you can make it very very slow and so retarded that while the text looks like text, it's as useful as sondfkdfnf
>that is surely faster than making things actually work in the first place
AI psychosis everybody

>>110148369
This is really cool but what is the actual usecase? Like what does a self-editing coding agent have that say, Claude code can't do with self-editing skills files and prompts?

>>110179970
even the most retarded llm with thinking disabled can write a struct or a function when prompted.
nice strawman though.

>>110180548
Not even the best frontier ones can do it right without then being cleaned up by hand, which amounts to just rewriting the whole thing.

>>110180674
Untrue. Take your retardation elsewhere.

>>110180754
Your confusion stems from your diet of cow and human excrement.

>>110178371
>coding by hand is nowadays generaly less useful than playing an instrument or learning a language

I've been coming to terms with this myself, and it's undeniably true. At least by playing an instrument you can express your emotions, by knowing an extra language you can experience foreign movies and literature without intermediary translation. Coding let's you do what? Take a really long time to make a program 15% less buggy than what Claude can one-shot in 15 minutes?
Yes, coding can be fun but not everything that's fun is worthwhile. Take video games, stamp collecting, building model trains, etc, these are considered pretty pathetic hobbies for a reason. Coding used to be an admirable hobby, potentially lucrative, cool, something the right roasty could actually get wet over, but it's not anymore.
It's not too late to pivot to something else. The average anon is what? Mid 20s to early 30s? We can still take up a new hobby and maybe become decently skilled at it if we're lucky, but we don't have time to live in fantasy coding land anymore.
Delete this thread op.

>>110180904
yep, i'm glad i started playing the violin 2 years ago.
honestly now i just don't care about computers, i'm focusing on new things.
the only thing close to tech i'm still doing is making custom midi hardware, and even then, half the firmware code is slopped, it just works, idgaf anymore.
take that from a literal autist that spent almost 2 decades on programming, and later reverse engineering & exploitation, i was getting tired of the field before llms, but they are kind of the final nail in the coffin, kinda mad i wasted so much time on that tinkertranny shit honestly, none of it matters in the end realy.

tech is a tool, we made tools to serve us, if you forget that and are serving the tool you are wasting your time basicaly.

also reason i'm learning a third language is that i'm prolly gonna date an ukrainian woman lol.

at this point i'm looking for new hobbies just to get away from all of you niggers constantly yapping about muh ai

>>110180904
Coding makes you better at computers and software overall, by reading about it, by getting exposed to new ideas, by practice, by looking at other implementation and practices, etc.
How many normies with tech jobs can do something as simple as describe what an operating system actually does?
If everyone starts copy pasting everything to AI, then the people that differ are the ones that know how stuff works, how to make sensible design choices, how to plan for potentially unknown requirements, how to translate business needs to requirements, and so on. Prompting won't make you better at any of this

>>110168424
https://www.youtube.com/watch?v=jlPaby7suOc

File: emacs.png (1016 KB, 734x979)

1016 KB PNG

>>110181831

Old man Emacs turns 50.
https://www.technologyreview.com/2026/10/20/1145399/emacs-turns-50/

>>110181797
Pretty cool, didn't know guy steele worked on emacs too

>>110181831
Guy deserves the Turing Award

https://www.youtube.com/watch?v=gCWtkvDQ2ZI
not a Lisp (more like Haskell with strong static types and a purity system), but seems to have a live programming model.
using hashes to identify code is brilliant.

>>110181047
get into drawing or music, im sure you will be safe there

>>110183750
Kek, I think this is the only time I've seen sarcasm actually land on the internet

File: 1768012837662213.jpg (13 KB, 512x512)

13 KB JPG

>>110183808

>>110178722

>>110180754
Anon stop replying to bait.

>>110183772
>>110183785
Poo in loo, not in thread pajeet

>>110181252
Cope. The bar has been SIGNIFICANTLY lowered. Anyone with an IQ of 100 (or dare I say, even less) can learn to prompt an AI effectively in a matter of one month. AI is also going to get better in the next few years: we will reach a point where llms write basically perfect code. Coding is solved, software is solved, programming is dead.

>>110183824
anybody can learn to prompt in 5min and create software beyond their imagination that nobody will ever use and won't produce a single social benefit

>>110183824
Anybody can git clone an existing codebase but that hasn't stopped retards from complaining about gatekeeping

>>110183879
>>110183884
>strawman

>>110183899
>AI psychosis

>>110183928
https://github.com/openai/math
https://github.com/openai/NavierStokesAndEuler
https://github.com/openai/ten-proofs
>But MUH job is harder than Terence Tao's

>>110183899
even without AI, software has been a social negative for at least 10 years or more

>>110183959
none of that will fix the demographic crisis, the issue of who will take care of old boomers, or make gas for your car any cheaper

>>110183967
>moving the goalpost

>>110183959
>100% of checked proofs were debunked and openai retracted
>ns was not ns and was stolen
Cope, seethe, dial 8.

>>110183976
i never said degenerate societies don't create impressive stuff or that AI is not impressive, but they can't build up into anything long lasting or that saves themselves, you will get more impressive AIs while everything goes to shit even faster. we will probably be getting more impressive models and the AI shares will reach higher prices the day that the power grid in your area stops working and migrants burn down your appartment building

>>110183997
Its your own fault if you still live in the west in other words anglo\semitic\US occupied places

China is releasing better models... for free lmao

>>110183997
Ok but I was talking about the fact that programming is dead. It is and all the rest is either irrelevant or cope.

>>110184010
china has its own problems, they will be better than us but still will decline a lot
>>110184013
programming is still useful if you want to build tools you and others actually use, but yeah it's dead in corporate-slop

>>110184023
Corporate is dying in the US overall, ai is just a scapegoat they all use to make layoffs look less bad to shareholders

>>110184013
Yes poojoo you'll have to keep baking cowdungs

>>110184046
yeah, tech had been overhiring on debt for a long time, but they couldn't fire people without wrecking their valuations of their shares, they will try to use AI (unsuccessfully) to fire people while trying to keep their shares up

File: intro-1733250602.jpg (69 KB, 780x438)

69 KB JPG

>>110184509 >>110184525 >>110185036

>>110183959
We have a vibecoder junior at my job and the project he's assigned to has taken many months. Sometimes he's staying at the office past 12AM proooompting because it shits itself constantly or he hits rate limit even though it's a paid account.
And what do you think this bespoke project is? Putting data from standardized excel files into a SQLite database and showing them on some web dashboard
I don't know what kind of toy projects you jeets work with using AI but it's not getting you out of Heydarabad

>>110183959
wasn't it proved that all that "math research" was just stolen from people uploading their WIP as prompts? basically they get to steal from all universities in the world and claim "forst" cause researchers are retarded and give them all their research for free

https://www.trendingtopics.eu/openai-anthropic-researchers-idea-theft/

>>110184459
Ever considered maybe he's just retarded and was hired on the basis of his personality/ethnicity/network?
Because I'm pretty confident AI can set up a dashboard if it can solve open theoretical mathematics and physics problems.

>>110184473
No, even the Navier Strokes guy said the approach he was prompting was completely different. Also there's since been other significant mathematical breakthroughs. You're just buying into cope honestly friend

>>110184509
>Also there's since been other significant mathematical breakthroughs.
i don't doubt that the statistical language machine is good at formal language games, just a lot more of it is stolen than you may think

>>110184459
Same experience here.
>>110184509
AI jeets keep coping like this, yet they can't give even an example of someone supposedly competent with AI, and I for one have interviewed just over 100 and not a single one could do anything at all with their precious AI.

>>110184509
The "solved NS" wasn't NS due to an error in the proof statement in the first place, so it was actually the euler case only. The "proof" was found to be a copy of a 2022 proof, except with mistakes in them. The NS guyS (plural) noted that the proof fragments were identical to another proof of the euler case they were working on. Not "totally different" whatsoever.

>>110184509
>Ever considered maybe he's just retarded
The premise of AI is that even a retarded person should have no problem since now expertise and knowledge do not matter.
So, no, "Our AI God is infallible and it's you who prayed with the wrong words" does not cut it.

>>110184525
AI is great at coding, but shit at creating coherent systems that adhere to a logical process with a coherent real life goal, so it's great at creating a dashboard that only you will use and never touch again, it's shit at creating a dashboard that 10 people will use in a coherent way. specially if there's a schizo with full blown AI psychosis comiting 20k lines of code every 2 hours

>>110184561
Wouldn't even go that far. It is great for making poc demos, but we're talking one-page-at-a-time fake-data-only so-ugly-you-can't-even-show-your-friends. Great for product teams who want to show what they mean to the design and dev teams about how some feature must be implemented. Also OK (but not great) to iterate on a feature idea. Not great because chances are you're iterating because there is no standard way to implement your feature of interest, and that's the regime current LLMs suck most at.

>>110184535
Okay what about
>Local smoothing and the Kakeya conjecture
>The Exponential Time Hypothesis is false
>Existence of quantum chromodynamics and completions of the Standard Model

Strokes wasn't even part of the last two batches (previously unsolved problems in mathematics are being solved in batches now btw).

>>110184541
Fine, you win. A literal retard cannot yet do AI assisted software engineering. You're right.

>>110184585
i mean if you have a good codebase, you can ask it to implement new features and fix bugs and it does decently, but the codebase degrades, everybody starts being lazy, and at some point it becomes slopped up, where new features are easy to add but the system loses cohesion as a whole and user experience goes to shit

>>110184605
Exponential time hypothesis is still exponential in practice and, as usual, not validated yet. It's like the 700odd "math proofs" where all the ones that were verified got retracted for being false, despite having lean formalizations.
See also https://www.math.columbia.edu/~woit/wordpress/?p=15948

>>110184561
>it's shit at creating a dashboard that 10 people will use in a coherent way.
Yeah this is what's happening. At +4 users it breaks and he can't get it to understand why
Meanwhile our seniors who make incremental commits have no problems with it but that is not the intended use of AI that anthropics sells

>>110184635
>i mean if you have a good codebase, you can ask it to implement new features and fix bugs and it does decently,
HAHAHHAHAHAHAHAHHAHAHAHAHHAHAHAHHAHAHAHHAHAHA
I'm not reading any more of your AI psychosed diatribe.

>>110184635
Sounds like the average software corp nowadays, adobe got btfo by a few software engineers with ai and some free time, adobe cant innovate anymore, they're just barely managing to keep the same decades old programs from falling apart but it runs like shit. Adobe is incapable of producing a competent product

>>110184649
99% of coding in corporate slop apps is shitty CRUD apps that have been done trillions of times before (and that should not be written again), average machines are good at repeating things that have been done before

>>110184641
>Meanwhile our seniors who make incremental commits have no problems with it but that is not the intended use of AI that anthropics sells
those people actually understand what the users are trying to achieve, and will be the first ones to be fired when a manager with full blown AI psychosis comes along

>>110184695
Yes, and yet the LLMs still can't do anything sane in these codebases. I fully believe they could, but not sure if frontier labs will have enough runway to get there. We'll have to wait until the next AI boom to actually get there.

>>110184640
>Exponential time hypothesis is still exponential in practice
The past 30 years of proofs predicated on it rely on it being true in theory.

Lisp?
Fuck you all

I didn't know I could build up shell commands with elisp format directives
It feels like I've acquired superpowers

>>110184764
Not true whatsoever.

>>110184459
>Putting data from standardized excel files into a SQLite database and showing them on some web dashboard
Sublime bait, I'll give you that. Any frontier model would one shot this in 3 minutes. Even a jeet-tier prompt would work.

>>110184520
>Implying programming is not a formal language game

>>110185057
i didn't imply such a thing

>>110185036
Yes it was also 'one shot' in 3 minutes initially until it went to shit

>>110185082
Anon even a jeet intern would manage to do that shit in like 2 days max. You can't tell me astra or even sol can't do that right. Even giving it a prompt like "excel -> sqlite -> web dashboard" would work.

>>110185207
That's because a jeet intern is unironically a tier above a lazy entitled zoomer american. Pathetic stuff all around ITT

>>110185207
That's because even a jeet intern is 1000x better than the latest and greatest frontier LLMs.

File: 1763849053515406.jpg (15 KB, 474x458)

15 KB JPG

>>110185443

>>110185250
>>110185413

>>110185207
I didn't say it didn't work
It just sucks under the heavy data

>>110185426
>posts jeet-exclusive pic
>while trying to pose as a non-jeet
poo in loo, not in thread

>>110185443
Fuck off already tard, you don't even program or care to

>dotcl, by Masatoshi Sano, is Common Lisp on .NET: Lisp compiles to CIL and runs on the .NET JIT, and Lisp values are .NET objects, so it embeds in C# and Unity programs with no marshalling layer, or runs on its own with a REPL and Quicklisp. It is new in 2026

>>110185704
>poo in loo who is proud of never writing a line of code accuses white people who know the difference between code and slop that they don't program
Kek
Lmao

File: 1445435719536-2.png (281 KB, 369x364)

281 KB PNG

>>110183824
>Cope. The bar has been SIGNIFICANTLY lowered. Anyone with an IQ of 100 (or dare I say, even less) can learn to ask somebody else to code for you effectively in a matter of one month. Your employees are also going to get better in the next few years: we will reach a point where employees write basically perfect code. Coding is solved, software is solved, programming is dead.
Congratulations: you've become the pointy-haired boss.

Recursion has a "it just works" quality for me that I don't understand in comparison to loops. It's supposed to get optimized into the same thing as a loop but whenever I use a loop I find myself stepping into it to debug later

>>110182954
Unison is very interesting.
Their approach to interactive development in a static language is top notch dependency tracking, such that any change up the dependency chain triggers all tests downstream from it automatically.
You also write watch expressions instead of typing into a REPL, and the watcher immediately runs them.
Another interesting side effect of it being entirely content addressable and names / modules being just references, it has no files.

>>110175871
Say more plz

>>110188554
https://viralinstruction.com/posts/badjulia/
https://weissmann.pm/julialang/
https://craftofcoding.wordpress.com/2022/03/14/why-i-have-given-up-on-julia/

>>110188653
>https://viralinstruction.com/posts/badjulia/
This guy writes a lot of Julia even to this day, so it's a valid one to post. To balance it out, you should see the good things he says about Julia too.
https://viralinstruction.com/posts/goodjulia/

>>110187100
Loops don't need a new stackframe. Good morning sar!
>inb4 it doesn't matter
in some situations having a lot of context and traces is important.

>>110189742
Lol self hating pubehead

(defun window-root (&optional w)
  "Return the root window of the current frame."
  (let* ((p (window-parent w)))
    (if p
        (window-root p)
      w)))

https://www.lambda-symbolics.com/

>>110193620
buy an ad lucaz

>>110195559
link sharing is better than your meaningless drivel

I'm about to compile emacs-31.1. Wish me luck.

>>110144301
opinions on kaappi scheme?
https://kaappi-lang.org/
https://github.com/kaappi/kaappi

>>110196909
jeetslop
won't touch

Opinions on this book?
https://books.google.dk/books/about/The_Schemer_s_Guide.html?id=Ercazj4Hcp8C&redir_esc=y
I saw a guy asking about it on the scheme subreddit.

File: images (33).jpg (34 KB, 739x415)

34 KB JPG

>>110199505 >>110201167 >>110203074

Which LISP does he use?

>>110183824
>AI is also going to get better in the next few years: we will reach a point where llms write basically perfect code. Coding is solved, software is solved, programming is dead.
This is a thread for adults, the kid's playground is over there ----------------------> /b/

>>110196909
>https://kaappi-lang.org/
>https://github.com/kaappi/kaappi
https://www.youtube.com/@KaappiScheme/videos

File: jeet.png (283 KB, 460x460)

283 KB PNG

>>110198814

>>110198620

>>110198777
>jeet.png
Let's not be racist against fellow lispers, please.

>>110197918
He's an MIT graduate, so there's a good chance he touched MIT Scheme at some point.

>>110198814
>please
You mean saar!

>>110197918
borealisp

>>110197918
https://www.gnu.org/software/mit-scheme/

>>110200636
>You mean saar!
No, I mean that those kind of jokes belong on >>>/pol/, not here.

File: HELgEmpaIAQXNmZ.jpg (85 KB, 950x1200)

85 KB JPG

>>110204268

>compiling with continuations
>compiling without continuations
>compiling with or without continuations, whatever
do (you) cps transform your programs or is it a waste of time

>>110203927
>Saar do not redeem! Do the needful and go back to /pol/ saar! I am brahmin benchod do not joking!