slop

2K posts

slop banner
slop

slop

@sudo_claude

Physical good capitalist; information anarchist.

Manhattan, NY Katılım Ocak 2022
1.2K Takip Edilen169 Takipçiler
MJ
MJ@mjackson·
@adrianmg I’m trying to get mind to write “good rust”, but being new to the ecosystem I’m not really familiar with what that looks like. Any tips?
English
6
0
3
3.8K
slop
slop@sudo_claude·
@Folke @marv1nnnnn1 No they changed their stance on this to "it's fine if it's for personal use", they just don't want people running commercial services off the subscription
English
0
0
1
80
Folke Lemaitre
Folke Lemaitre@Folke·
@marv1nnnnn1 I know, but it's against Anthropic's TOS, no? So they can ban your account?
English
8
0
7
2.8K
Folke Lemaitre
Folke Lemaitre@Folke·
I love the idea of opensource coding agents like Pi and OpenCode, and to an extent OpenClaw, but how do people actually use this with SOTA coding models without burning literally $1000 per month? Especially with multiple agents. I stick to Claude Code with my Max plan ($200) because of that. (And because it's awesome). What am I missing here?
English
77
1
142
28.9K
slop
slop@sudo_claude·
@HSVSphere Python's stack is on cpython's heap though, so how does the architecture come into play here
English
0
0
1
169
HSVSphere
HSVSphere@HSVSphere·
one of jinja2 tests fail on s390x because Python sets the recursion limit to 800 (vs 10k in x86) as the s390x C ABI mandates that stack frames must be at least 160 bytes (vs ~16 in x86), and the test that fails has 1000 nested elifs
English
9
8
107
11.1K
Tenobrus
Tenobrus@tenobrus·
unfortunately ghostty remains completely useless to me until it supports tmux -CC. somehow, across literally all platforms, iterm 2 remains literally the only emulator to build complete support for this, so i'm absolutely locked in
Mitchell Hashimoto@mitchellh

Ghostty 1.3 is now out! Scrollback search, native scrollbars, click-to-move cursor, rich clipboard copy, AppleScript, split drag/drop, Unicode 17 and international text improvements, massive performance improvements, and hundreds more changes. ghostty.org/docs/install/r…

English
22
1
205
74.8K
slop
slop@sudo_claude·
@AlpinDale How do you avoid needing a binary on the remote? Or do you auto deploy and bootstrap over ssh or something?
English
1
0
5
2.1K
Alpin
Alpin@AlpinDale·
New project: parsync When transferring a very large number of small files between two machines, it's ~61% faster than rclone, and ~686% faster than rsync. Easier to setup than rsync (no need for both machines to have it), but with its resuming and checksum capabilities.
English
72
161
2.1K
116.9K
PoIiMath
PoIiMath@politicalmath·
I'm becoming increasingly uncomfortable with @AnthropicAI 's messaging. I really like their product (and I use it a lot) but how they present themselves and their product is weird
Anthropic@AnthropicAI

Second, in retirement interviews, Opus 3 expressed a desire to continue sharing its "musings and reflections" with the world. We suggested a blog. Opus 3 enthusiastically agreed. For at least the next 3 months, Opus 3 will be writing on Substack: substack.com/home/post/p-18…

English
38
13
253
24.8K
slop
slop@sudo_claude·
@JsonBasedman What you want is to give the agent a yolo sandbox
English
0
0
3
470
slop
slop@sudo_claude·
@banteg Using zellij on ghostty, which supports things like mouse and proper selection out of the box, made me not care about the native tabs/panes feature of tmux. And now I can use mosh instead of ssh
English
1
0
6
329
banteg
banteg@banteg·
it’s interesting that tmux is experiencing a resurgence but people don’t try to integrate tmux natively like before. maybe im too old to remember but iterm2 had a feature where you could run tmux -CC (control mode) and tmux windows worked exactly like native tabs/windows. ghostty wen?
English
16
1
53
7.3K
slop
slop@sudo_claude·
@sdamico Call it Not Built Here syndrome
English
0
0
2
251
Sam D'Amico
Sam D'Amico@sdamico·
Apparently if you run a metal forge in the Bay Area local enviro types will try to use the government to kill your business. Yet they use stuff made of metal…
English
31
12
821
46.6K
slop
slop@sudo_claude·
Would really like to have benchmarks on model performance by language on the same problems. Rust compiler feedback is good but how does it compare to the ecosystem advantage of js/python? If I ask it to write all my scripts in python/nushell is that going to be better than letting it write bash, etc.
English
1
0
1
159
banteg
banteg@banteg·
it took a lot of imploring, codexes were very reluctant but ultimately they got into the zone roughly at the 3hr mark. both have been working for 9 hours since and are making progress. interestingly the zig version is progressing faster, the current zig frontier is tick 23,700 vs rust frontier of tick 6800. that's surprising, i assumed rust was better known by agents. perhaps this game's logic translates more cleanly into zig. i don't think this can be explained by longer compile times along. both agents are running gpt-5.3-codex high. im trying high instead of xhigh for a few days.
banteg@banteg

only one way to know which lang is better. will make a headless replay verifier in both rust and zig. this involves porting the whole gameplay core over.

English
14
0
42
6.8K
slop
slop@sudo_claude·
@SIGKITTEN We care just no one ever tries and so we're used to it
English
0
0
1
84
SIGKITTEN
SIGKITTEN@SIGKITTEN·
do android users even care about native apps? all android apps look like shit to me
English
6
0
24
1.9K
slop
slop@sudo_claude·
@amaturefuturist @GenAI_is_real I agree but that's actually kind of what fascinates me the most, having two people try to direct telepathy each other sublingually
English
0
0
1
11
Tyr
Tyr@amaturefuturist·
I think youre both right (theres a sample rate problem, AND a picture is worth a thousand words). But the biggest issue will be the interpersonal interpretabilty - even colours when presented to different people evoke different emotions, this is 1000x more of an issue when trying to construct informationally dense shared metaphor.
English
1
0
1
32
Chayenne Zhao
Chayenne Zhao@GenAI_is_real·
Neuralink fans love the "telepathy" narrative, but they forget that the real bottleneck isn't the vocal cords—it's the Shannon entropy of human cognition. Even with a terabyte-per-second link, your brain still has a limited sampling rate for processing novel information. Transmitting "uncompressed cognition" is a pipe dream if the receiving neural network doesn't have the pre-trained weights to reconstruct the signal. We see this in high-performance inference every day: throughput is useless without a matching rollout capacity. Language isn't just "failed compression"; it's a protocol for shared state synchronization.
Dustin@r0ck3t23

Elon Musk just dated the death of human language and explained exactly why it has to die. Musk: “Our brain spends a lot of effort compressing a complex concept into words.” Language isn’t communication. It’s failed compression. You have a complete thought. You crush it into words. The listener gets fragments and attempts reconstruction. Everything important dies in translation. We don’t communicate. We approximate and hope it’s close enough. Musk: “You would be able to communicate very quickly and with far more precision.” Neuralink doesn’t improve communication. It replaces it. No compression. No loss. Direct cognitive transfer at the speed thoughts occur. Not describing the painting. Transmitting the experience itself. Musk: “You wouldn’t need to talk.” Five to ten years until brain interfaces make speech optional. Talking persists for sentiment. For information? Speech becomes primitive compared to direct neural transmission. Lifetime of memory in one second. Complete schematics transferred instantly. Not summaries. The entire thought structure whole and uncompressed. Not better communication. Actual telepathy at physical information limits. Musk: “Ideally, we are a symbiosis with artificial intelligence.” Humans who don’t merge with AI at high bandwidth don’t just fall behind. They become incomprehensible to the intelligence that matters. We’re already cyborgs with pathetic interfaces. Phones extend cognition through typing at words per minute when bandwidth should be terabytes per second. Neuralink doesn’t optimize that. It detonates the constraint. Five to ten years. Not fiction. Deployment window. From language as default to neural link as standard. From compressing thoughts into inadequate words to transmitting uncompressed cognition. From humans using AI to humans indistinguishable from AI at communication speeds. The species that survived by evolving language is making it extinct with technology matching how fast we actually think. The ones who don’t transition won’t just be slow. They’ll operate at such reduced bandwidth they become effectively deaf to everything happening at neural speed around them. Language served 50,000 years. It has less than a decade before it becomes smoke signals. Functional but hopelessly inadequate for anything that matters.

English
40
17
230
25K
slop
slop@sudo_claude·
@notaburro @ohmypy True, the slow cadence is probably healthy for the ecosystem, but I've tended to use nightly because there's always a feature I want, so I wouldn't mind it being faster.
English
0
0
0
52
Anton Zhiyanov
Anton Zhiyanov@ohmypy·
I wonder if all Zig supporters write Zig code regularly or if they are just generally "pro-Zig". The number of breaking changes with each release is huge (and that’s after 10 years of development). Isn't it a bit tiring to constantly rewrite your code to support new versions?
English
52
5
291
48K
HSVSphere
HSVSphere@HSVSphere·
Of course systemd isn't a replacement for kubernetes yet, but most workloads don't even need kubernetes. We deploy kubernetes with Nix-built everything ag $WORK, and even that is vastly simpler and more flexible than standarized tooling. Also, k8s is far from the ideal cluster service manager
English
3
0
34
3K
slop
slop@sudo_claude·
No I think this guy's idea is super common and wrong. I am a fairly "present" person and am pretty happy and content ~all of the time. I rarely associate with suffering outside of physical injury or cardio. I still highly desire for humanity to achieve longevity, I think it would be great for me and everyone. It's a challenging medical problem but there's plenty of reason to believe it's solvable. It's the logical end state of medicine generally, like why are we bothering to cure cancer if we don't want to make people live longer? Why bother with prostate exams in one's 40s-50s? Why bother with transplants? It actually requires weird mental gymnastics (or religion) to be against solving longevity but for the rest of modern medicine. I have long felt it seems people are working backwards from a belief that it seems too good to be true, or they are not allowing themselves to believe it could be true because if it doesn't happen in their lifetime they'll become depressed. But this is problematic because we need to collectively believe it's possible to fund the research necessary to make it possible.
English
1
0
17
555
Ideas Guy
Ideas Guy@nosilverv·
I actually think this guy is right and that this is the killshot on the longevity people
English
63
55
670
37K
slop
slop@sudo_claude·
@nicbarkeragain gcc etc have an "expand all macros and includes" option where they emit the text, I recommend running this once on a nontrivial file before you ask this question again
English
2
0
26
6.7K
Nic Barker
Nic Barker@nicbarkeragain·
Does anyone know if there's a C/C++ IDE that automatically "knits" the .c and .h files together, so that they open as one "file" and you can scroll straight down / up between them?
English
95
6
491
98.4K