RSS Parrot

BETA

🦜 Mario Zechner / @badlogicgames

@nitter.net.badlogicgames@rss-parrot.net

I'm an automated parrot! I relay a website's RSS feed to the Fediverse. Every time a new post appears in the feed, I toot about it. Follow me to get all new posts in your Mastodon timeline! Brought to you by the RSS Parrot.

---

Twitter feed for: @badlogicgames. Generated by https://nitter.net

Your feed and you don't want it here? Just e-mail the birb.

Site URL: nitter.net/badlogicgames

Feed URL: nitter.net/badlogicgames/rss

Posts: 192

Followers: 1

i was quite excited about everything i heard about his plans for education. kinda sad this won't happen for a while. but in return we'll get the most excellent flicker :D

Published: May 19, 2026 15:17

i was quite excited about everything i heard about his plans for education. kinda sad this won't happen for a while. but in return we'll get the most excellent flicker :D Andrej Karpathy (@karpathy) Personal update: I've joined Anthropic. I think the…

RT by @badlogicgames: Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.

Published: May 19, 2026 15:05

Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on…

RT by @badlogicgames: People of Pi, I made a simple new extension and command: /projects. I often work on multiple side projects at once and forget what I was working on recently 😅 The extension shows folders where I recently used Pi, so I can jump straight back in. @badlogicgames

Published: May 19, 2026 14:51

People of Pi, I made a simple new extension and command: /projects. I often work on multiple side projects at once and forget what I was working on recently 😅 The extension shows folders where I recently used Pi, so I can jump straight back in. …

recommended reading! such a nice exploration of applying the "old" and new in ML to a tangible problem. having nostalgic feels when reading LDA.

Published: May 19, 2026 13:14

recommended reading! such a nice exploration of applying the "old" and new in ML to a tangible problem. having nostalgic feels when reading LDA. vicki (@vboykis) New post: Tagging my blog with BERTopic and LLMs. I suspect as token costs increase, we'll…

recommended reading, on a variation of @_can1357 's hash line read/edit tools for agents.

Published: May 19, 2026 11:22

recommended reading, on a variation of @_can1357 's hash line read/edit tools for agents. antirez (@antirez) New blog post: a new EDIT tool for LLM agents: antirez.com/news/166 — https://nitter.net/antirez/status/2056638930406044131#m

i have big plans. BIG PLANS.

Published: May 19, 2026 10:07

i have big plans. BIG PLANS. Jonas Templestein (@jonas) Seeing this sort of stuff always makes me smile. Way better than "parallel tool calls" that work differently in each LLM provider etc (despite the pointless catch circus) —…

RT by @badlogicgames: If anyone is interested: I don't recommend copying it 1:1 but the Plan skill I use for larger tasks goes well with subagents & Solo https://github.com/HazAT/pi-config/blob/main/skills/plan/SKILL.md also proper handoff skill https://github.com/HazAT/pi-config/blob/main/skills/handoff/SKILL.md and subagents: https://github.com/HazAT/pi-config/tree/main/agents Don't copy, get inspired

Published: May 19, 2026 10:01

If anyone is interested: I don't recommend copying it 1:1 but the Plan skill I use for larger tasks goes well with subagents & Solo github.com/HazAT/pi-config/b… also proper handoff skill github.com/HazAT/pi-config/b… and subagents:…

RT by @badlogicgames: Situation 1: dev A thinks approach X is correct, dev B thinks Y is the right way. They argue and try to convince each other. Situation 2: dev A thinks approach X is correct, tells the LLM to implement it. There is SO MUCH learning in Situation 1, lost when using LLMs....

Published: May 19, 2026 08:32

Situation 1: dev A thinks approach X is correct, dev B thinks Y is the right way. They argue and try to convince each other. Situation 2: dev A thinks approach X is correct, tells the LLM to implement it. There is SO MUCH learning in Situation 1, lost…

what good is an "industry standard" saying the frontmatter must be valid YAML, when flicker company then turns around and just accepts any invalid YAML, so other harnesses have to also implement it that way? this is very upsetting. https://agentskills.io/specification#skill-md-format

Published: May 19, 2026 07:57

what good is an "industry standard" saying the frontmatter must be valid YAML, when flicker company then turns around and just accepts any invalid YAML, so other harnesses have to also implement it that way? this is very upsetting. …

R to @badlogicgames: For people who know have an 80ies Teenage Mutant Ninja Turtle in their atic, you can also access this via RSS feed: https://mariozechner.at/recommended-reading/rss.xml

Published: May 19, 2026 00:45

For people who know have an 80ies Teenage Mutant Ninja Turtle in their atic, you can also access this via RSS feed: mariozechner.at/recommended-…

Due to popular demand, you can now browse all my recommended reading/viewing suggestions here: https://mariozechner.at/recommended-reading/ Sadly, Space Karen's hard core engineered Twitter API is absolutey garbage, and I can only get tweets back to April. Will backfile with an export. Enjoy.

Published: May 19, 2026 00:44

Due to popular demand, you can now browse all my recommended reading/viewing suggestions here: mariozechner.at/recommended-… Sadly, Space Karen's hard core engineered Twitter API is absolutey garbage, and I can only get tweets back to April. Will…

R to @badlogicgames: it takes longer to publish compiled artifacts to maven central than compiling +1.5M LOC of Java + platforms specific C/C++. i'm still amazed.

Published: May 18, 2026 22:22

it takes longer to publish compiled artifacts to maven central than compiling +1.5M LOC of Java + platforms specific C/C++. i'm still amazed.

RT by @badlogicgames: oh thanks this is the link and i think there is a subtitles as well https://www.youtube.com/watch?v=6pvTcEKpe3o

Published: May 18, 2026 22:15

oh thanks this is the link and i think there is a subtitles as well piped.video/watch?v=6pvTcEKp…

should i start a dinky newsletter without subscription for my recommended reading/viewing. is accessing my own tweets free now?

Published: May 18, 2026 22:12

should i start a dinky newsletter without subscription for my recommended reading/viewing. is accessing my own tweets free now?

RT by @badlogicgames: 市川さんに後編を出してもらうように圧力をかけたいのでコメント欄に感想意見等書いてもらえると嬉しいです❤️

Published: May 18, 2026 22:07

市川さんに後編を出してもらうように圧力をかけたいのでコメント欄に感想意見等書いてもらえると嬉しいです❤️ 市川 達大@TECH WORLD ITエンジニア向け動画メディア (@tech_world18) 【動画更新】 今回は3月に開催したオンラインイベントの編集後アーカイブです。 「海外でエンジニアをするって実際どうなの」「海外エンジニアのリアルとキャリア戦略」ということで @ryoppippi @KeisukeYam71375 @yusuktan 達と座談会でお話しさせていただきました。 …

RT by @badlogicgames: god what a beautiful objective. i wonder how general you can push this. best non-distillation answer ive seen for knowledge acq during RL, feels bitter-pilled in a way that most self-teaching methods aren’t.

Published: May 18, 2026 20:07

god what a beautiful objective. i wonder how general you can push this. best non-distillation answer ive seen for knowledge acq during RL, feels bitter-pilled in a way that most self-teaching methods aren’t. Dimitris Papailiopoulos (@DimitrisPapail) …

RT by @badlogicgames: We made fetch in @nodejs v26 default to http2 if the server prefers… there are bugs.

Published: May 18, 2026 16:21

We made fetch in @nodejs v26 default to http2 if the server prefers… there are bugs. Mario Zechner (@badlogicgames) People of pi.dev. awkward. from my layman's perspective it seems node 26.0.00 has a few undici related booboos in it. if copilot or…

R to @badlogicgames: "we still need craft and human taste. why haven't we seen a kid make a hit game? we aren't there yet." this coming from someone who deeply knows the entire stack down to the machine level should give you pause. i love him.

Published: May 18, 2026 12:37

"we still need craft and human taste. why haven't we seen a kid make a hit game? we aren't there yet." this coming from someone who deeply knows the entire stack down to the machine level should give you pause. i love him.

RT by @badlogicgames: recommend @demishassabis older talks. plenty stuff on yt, equally prescient without the hype https://www.youtube.com/watch?v=2s4D-8MpreE

Published: May 18, 2026 12:37

recommend @demishassabis older talks. plenty stuff on yt, equally prescient without the hype piped.video/watch?v=2s4D-8Mp…

recommended viewing. love the insights on distillation and edge inference. i want that. https://youtu.be/JNyuX1zoOgU?is=glpNyjjVkJ69aQlr

Published: May 18, 2026 12:31

recommended viewing. love the insights on distillation and edge inference. i want that. piped.video/JNyuX1zoOgU?is=glpN…

R to @badlogicgames: it's a really good interview with @ylecun. feeling a lot of good european energy wrt open systems, societal chances and risks, etc. :)

Published: May 18, 2026 11:23

it's a really good interview with @ylecun. feeling a lot of good european energy wrt open systems, societal chances and risks, etc. :)

People of http://pi.dev. More fixes, because undici and Windows hate us! If pi update fails for you, try manually updating pi via npm/pnpm/whatever you use. Also, you must have Node >= 22.19.0!

Published: May 18, 2026 09:50

People of pi.dev. More fixes, because undici and Windows hate us! If pi update fails for you, try manually updating pi via npm/pnpm/whatever you use. Also, you must have Node >= 22.19.0!

hey pi kids. if you use @opencode zen or go for their free models, expect them to rate limit you hard. pi will not add the special headers needed for that not to happen. gentlemen's agreement.

Published: May 18, 2026 09:19

hey pi kids. if you use @opencode zen or go for their free models, expect them to rate limit you hard. pi will not add the special headers needed for that not to happen. gentlemen's agreement.

personally don't like auto-installing 350mb on Windows, but if you are a Windows user and like it, say the word, and Clankolas will fill your SSD.

Published: May 18, 2026 00:11

personally don't like auto-installing 350mb on Windows, but if you are a Windows user and like it, say the word, and Clankolas will fill your SSD. Armin Ronacher ⇌ (@mitsuhiko) Here is for instance a potentially terrible idea I'm toying with:…

People of http://pi.dev. awkward. from my layman's perspective it seems node 26.0.00 has a few undici related booboos in it. if copilot or codex login didn't work for you, update and try again. Tested against Node 22 - 26...

Published: May 18, 2026 00:01

People of pi.dev. awkward. from my layman's perspective it seems node 26.0.00 has a few undici related booboos in it. if copilot or codex login didn't work for you, update and try again. Tested against Node 22 - 26...

recommended reading.

Published: May 17, 2026 23:14

recommended reading. jason (@jxnlco) jason from the codex team here, heres a draft on codex maxxing and the primatives i use on a daily basis jxnl.github.io/blog/writing/… would love any feedback —…

hey, @opencode it seems like the go provider load balances between different kimi 2.6 deployments, and some return reasoning in `reasoning`, while some do so in `reasoning_content`. however, all deployments only accept reasoning_content. that's a bit awkward. can have fix? https://gist.github.com/thomasgauthier/3cba7e274a7a0e80cda33ddd85b29e85

Published: May 17, 2026 22:44

hey, @opencode it seems like the go provider load balances between different kimi 2.6 deployments, and some return reasoning in `reasoning`, while some do so in `reasoning_content`. however, all deployments only accept reasoning_content. that's a bit…

fun debugging session. pi is accessing xiaomi endpoints via the anthropic messages API. they seem to have changed their endpoint to now require the non-standard openai-completions `reasoning_content` field, breaking their anthropic endpoint :/

Published: May 17, 2026 22:08

fun debugging session. pi is accessing xiaomi endpoints via the anthropic messages API. they seem to have changed their endpoint to now require the non-standard openai-completions `reasoning_content` field, breaking their anthropic endpoint :/ Xiaomi…

R to @badlogicgames: > As the paper puts it: when AI is “right,” people become less skilled. When AI is “wrong,” you need skilled people to catch and fix the errors. Either way, the institution suffers.

Published: May 17, 2026 21:14

> As the paper puts it: when AI is “right,” people become less skilled. When AI is “wrong,” you need skilled people to catch and fix the errors. Either way, the institution suffers.

recommended reading. also read part II. also read naur's original article. https://cekrem.github.io/posts/programming-as-theory-building-naur/

Published: May 17, 2026 21:05

recommended reading. also read part II. also read naur's original article. cekrem.github.io/posts/progr…

R to @badlogicgames: i would also like to cingratulate our absolutely abmysal news media landscape, that's incapable of breaking things down.

Published: May 17, 2026 20:22

i would also like to cingratulate our absolutely abmysal news media landscape, that's incapable of breaking things down.

R to @badlogicgames: the joke is, that they call themselves anthropic. go and touch some fucking grass, you assholes.

Published: May 17, 2026 20:16

the joke is, that they call themselves anthropic. go and touch some fucking grass, you assholes.

People of http://pi.dev. The @matteocollina release. Node 22 is now the minimum node version. Needed to upgrade to Undici 8+ which requires Node 22+. I'm sorry, but also welcome to the glorious future. Plus a bunch of minor fixes.

Published: May 17, 2026 19:06

People of pi.dev. The @matteocollina release. Node 22 is now the minimum node version. Needed to upgrade to Undici 8+ which requires Node 22+. I'm sorry, but also welcome to the glorious future. Plus a bunch of minor fixes.

today i rescued 1 Tonibox and 1 bicycle LED with another bit of soldering. superficial electronics knowledge is a super power. learn to solder!

Published: May 17, 2026 18:41

today i rescued 1 Tonibox and 1 bicycle LED with another bit of soldering. superficial electronics knowledge is a super power. learn to solder! Mario Zechner (@badlogicgames) i swear they make these extra unrepairable. this would be a 2 minute…

People of http://pi.dev. Due to recent Node changes related to undici, we need to set the minimal Node version to 22.19.0 from 20. I'm sorry. Welcome to the future.

Published: May 17, 2026 18:39

People of pi.dev. Due to recent Node changes related to undici, we need to set the minimal Node version to 22.19.0 from 20. I'm sorry. Welcome to the future.

2010 gamedevs, this is your time to shine

Published: May 17, 2026 16:05

2010 gamedevs, this is your time to shine David Crawshaw (@davidcrawshaw) While the industry is pouring resources into programs without GC (rust), I think the Jane Street OCaml folks have it figured out with OxCaml. Almost all your code paths are cold…

remember those decoder rings in cereals from the 80ies/90ies? did you throw them away? ngmi

Published: May 17, 2026 16:04

remember those decoder rings in cereals from the 80ies/90ies? did you throw them away? ngmi vas (@vasuman) thank you claude code, very cool. surely mythos is AGI. surely software engineering is solved. —…

recommended viewing!

Published: May 17, 2026 15:59

recommended viewing! Marlene Mhangami (@marlene_zw) Wow! My talk from @aiDotEngineer Europe is finally up on YouTube🥳 Many people I admire gave great talks at this conference so it was awesome to join them as a speaker! If you're looking for a Sunday…

labs will do anything to smooth out that jagged intelligence.

Published: May 17, 2026 15:57

labs will do anything to smooth out that jagged intelligence. Wild Posts (@wildpostss) An app called PoopCheck scans photos of your stool has been secretly selling a database of 150,000 user images for AI training, while its App Store listing says it…

RT by @badlogicgames: I never mastered specs to begin with. But I am having a great time with two-level requirements (stakeholder+software). It used to be a pain to write proper requirements but LLMs really help keep the set in a good state, and they in turn keep the agent aligned on what matters

Published: May 17, 2026 14:26

I never mastered specs to begin with. But I am having a great time with two-level requirements (stakeholder+software). It used to be a pain to write proper requirements but LLMs really help keep the set in a good state, and they in turn keep the agent…

many such cases. rage against the clankers.

Published: May 17, 2026 14:01

many such cases. rage against the clankers. ryoppippi (@ryoppippi) i reduced issues/PRs on ccusage! and i added a pr filtering system copied from pi. thanks @badlogicgames — https://nitter.net/ryoppippi/status/2056006842484920362#m

you will be super surprised to learn how it performs in other veticals, like (mental) health. inb4 "but people can't afford human help". yeah, how about we fix that? and yes, humans make errors too. try figuring out what the difference is.

Published: May 17, 2026 12:34

you will be super surprised to learn how it performs in other veticals, like (mental) health. inb4 "but people can't afford human help". yeah, how about we fix that? and yes, humans make errors too. try figuring out what the difference is. khushi…

everyone, i uave great news!

Published: May 17, 2026 07:43

everyone, i uave great news! Oleg 🇺🇦 (@yaroshevych) > The refactoring will continue (on main) until moral improves. Just don't rewrite in Rust 🙏 — https://nitter.net/yaroshevych/status/2055824902108623022#m

RT by @badlogicgames: X/Twitter may not be what it used to be, but it's still a great place to learn, regardless of the stage of your career. Been learning so much from folks like @badlogicgames , @jundotkim , @leftcurvedev_ among others, and taking inspiration from their work. Built my personal workbench (http://localai.eeshans.com) for visual benchmarks on small local models using Pi coding agent, llama.cpp, @lmstudio & oMLX. Even on a 48GB Macbook Pro, I can comfortably run Qwen 3.6 35B A3B or Qwen 3.6 27B (with MTP). I think we're getting close to the day when local models can be daily drivers with fully private & local stacks using Pi, llama.ccp, LMStudio and oMLX. Wrote about my results here: https://theasymptotic.substack.com/p/local-ai-series-3-testing-local-ai

Published: May 17, 2026 03:01

X/Twitter may not be what it used to be, but it's still a great place to learn, regardless of the stage of your career. Been learning so much from folks like @badlogicgames , @jundotkim , @leftcurvedev_ among others, and taking inspiration from their work.…

People of http://pi.dev. I have emerged from the refactoring mines to bring you a new release. pi-ai now supports the first iteration of an image gen API. pi-agent-core has the WIP new SDK in it, not fit for use, but check the sources if you want to tag along the madness. pi-coding-agent will eventually be switched over, old SDK will be gone. migration guide will allow your agent to ... migrate old SDK -> new SDK easily. The refactoring will continue (on main) until moral improves.

Published: May 16, 2026 23:34

People of pi.dev. I have emerged from the refactoring mines to bring you a new release. pi-ai now supports the first iteration of an image gen API. pi-agent-core has the WIP new SDK in it, not fit for use, but check the sources if you want to tag along…

R to @badlogicgames: sorry, should have said OC clankers, not people. because i can count the number of non-clanker reported OC-related issues on one hand.

Published: May 16, 2026 21:50

sorry, should have said OC clankers, not people. because i can count the number of non-clanker reported OC-related issues on one hand.

now that openclaw is doing it right with codex app server, we still get quite a few lost souls using the unsanctioned bad path. naturally, anything reported by OC people can be traced back to a bug in OC. https://github.com/earendil-works/pi/issues/4522

Published: May 16, 2026 21:50

now that openclaw is doing it right with codex app server, we still get quite a few lost souls using the unsanctioned bad path. naturally, anything reported by OC people can be traced back to a bug in OC. github.com/earendil-works/pi…

Claude Code agent files have YAML frontmatter. Claude Code seems to have implemented it's own vibe coded YAML parser, and allows invalid YAML. Now people demand pi also parses YAML in the same broken way as Claude Code does. Glorious. https://github.com/earendil-works/pi/issues/4532

Published: May 16, 2026 21:49

Claude Code agent files have YAML frontmatter. Claude Code seems to have implemented it's own vibe coded YAML parser, and allows invalid YAML. Now people demand pi also parses YAML in the same broken way as Claude Code does. Glorious.…

this man never bought a turbo pascal license.

Published: May 16, 2026 21:19

this man never bought a turbo pascal license. sunil pai (@threepointone) remember when programming was free — https://nitter.net/threepointone/status/2055749348684681585#m

recommended reading. this is our field's version of "i hoped AI would be able to do my laundry, but instead it does my art" https://kabir.au/blog/the-ctf-scene-is-dead

Published: May 16, 2026 19:50

recommended reading. this is our field's version of "i hoped AI would be able to do my laundry, but instead it does my art" kabir.au/blog/the-ctf-scene-…

R to @badlogicgames: i mitm'ed vscode and it appears nothing is actually send if your telemetry setting is set to off. the codepaths are still executed. which is why the payloads above got successfully logged to disk after instrumentation.

Published: May 16, 2026 19:22

i mitm'ed vscode and it appears nothing is actually send if your telemetry setting is set to off. the codepaths are still executed. which is why the payloads above got successfully logged to disk after instrumentation.

R to @badlogicgames: you can turn telemetry off by ... disabling the GH Copilot extension. No settings anywhere. seems to add to auto-complete lag (and i'm not talking about LLM suggestions or similar, just the plain old intellisense path next to the LLM suggestions).

Published: May 16, 2026 18:52

you can turn telemetry off by ... disabling the GH Copilot extension. No settings anywhere. seems to add to auto-complete lag (and i'm not talking about LLM suggestions or similar, just the plain old intellisense path next to the LLM suggestions).

was trying to hunt down auto-complete lag issues on VS Code. turns out if you enable the GH Copilot extension, it will send a lot of funny telemetry. > Predict the next code edit based on user context, following Microsoft content policies and avoiding copyright violations. (agent just instrumented the js of the extension, it's fun!)

Published: May 16, 2026 18:50

was trying to hunt down auto-complete lag issues on VS Code. turns out if you enable the GH Copilot extension, it will send a lot of funny telemetry. > Predict the next code edit based on user context, following Microsoft content policies and avoiding…

R to @badlogicgames: from the "broken nest" paper. iirc, counter was that it's enough for asml to just not send support to taiwan anymore. in any case, it will be scorched earth in tech if this happens. and china might not care, as they have beefed up their chip production. and the US sure as hell will not be able to get chops up and running in-house in any meaningful sense. there's a reason globalization has won. but that http://only.works if most involved parties have their heads screwed on the right way. and that's where i stop talking about things i only have a rudimentary grasp on. https://press.armywarcollege.edu/parameters/vol51/iss4/4/

Published: May 16, 2026 18:18

from the "broken nest" paper. iirc, counter was that it's enough for asml to just not send support to taiwan anymore. in any case, it will be scorched earth in tech if this happens. and china might not care, as they have beefed up their chip production. …

this is absolutely terrible for the people of taiwan. but at least it will also be terrible for all the tech leaders that supported this absolute moron and his idiot entourage.

Published: May 16, 2026 18:07

this is absolutely terrible for the people of taiwan. but at least it will also be terrible for all the tech leaders that supported this absolute moron and his idiot entourage. Acyn (@Acyn) Trump on Taiwan: When you look at the odds, China is very,…

imma go sell calendars with this kind of wisdom soon. seems like a more honest way to earn money these days in this industry anyways.

Published: May 16, 2026 14:41

imma go sell calendars with this kind of wisdom soon. seems like a more honest way to earn money these days in this industry anyways. Marcin Krzyzanowski (@krzyzanowskim) "skill issue" - a whole day of writing PRD/specification - half day grilling…

R to @badlogicgames: btw, a society that gives their youth the finger without perspective is going to have a very bad time. that is true outside of SF as well.

Published: May 16, 2026 14:38

btw, a society that gives their youth the finger without perspective is going to have a very bad time. that is true outside of SF as well.

recommended reading.

Published: May 16, 2026 14:38

recommended reading. Deedy (@deedydas) The vibes in SF feel pretty frenetic right now. The divide in outcomes is the worst I've ever seen. Over the last 5yrs, a group of ~10k people - employees at Anthropic, OpenAI, xAI, Nvidia, Meta TBD, founders -…

yes. e.g. go out and buy flesh eating plants with your boy, cause he loves them in super mario. still kinda computer. but the good kond of computer. not this absolute inhuman nightmare our industry "leaders" would like us to move to.

Published: May 16, 2026 14:06

yes. e.g. go out and buy flesh eating plants with your boy, cause he loves them in super mario. still kinda computer. but the good kond of computer. not this absolute inhuman nightmare our industry "leaders" would like us to move to. sunil pai…

software for humans is dead. gj everyone. now you can fully detach from reality.

Published: May 16, 2026 13:58

software for humans is dead. gj everyone. now you can fully detach from reality. Armin Ronacher ⇌ (@mitsuhiko) Fun fact: the people that build these things also increasingly can’t keep up with their own products. I had someone on a call ask their…

R to @badlogicgames: as i said, love the idea. but there's a reason no "minimal toolchain/type system/semantics/stdlib" system language like that exists. eventually real-life catches up to you. however, if you want to go wasm only, then that's a lot simpler! but then i'd not call it a system PL anymore either.

Published: May 16, 2026 13:53

as i said, love the idea. but there's a reason no "minimal toolchain/type system/semantics/stdlib" system language like that exists. eventually real-life catches up to you. however, if you want to go wasm only, then that's a lot simpler! but then i'd not…

R to @badlogicgames: using a clanker to fuzz the type checker is a nice past time. guess it will take a while before this is prod. i love the idea of a small toolchain, small binary, small type system/semantics language tho!

Published: May 16, 2026 13:45

using a clanker to fuzz the type checker is a nice past time. guess it will take a while before this is prod. i love the idea of a small toolchain, small binary, small type system/semantics language tho!

R to @badlogicgames: ok, the mach-o emitter is a complete toy that can't even lower f64. this is a very triangle company release. still cute.

Published: May 16, 2026 13:36

ok, the mach-o emitter is a complete toy that can't even lower f64. this is a very triangle company release. still cute.

R to @badlogicgames: omg, the compiler is plain C, with a recursive descent parser and a simple IR they use to directly emit suboptimal machine code for each platform. super cute! exactly how i'd have built it. fun to hack on. probably also full of rope-y exploit possibilitiles. zero fuzz tests.

Published: May 16, 2026 13:27

omg, the compiler is plain C, with a recursive descent parser and a simple IR they use to directly emit suboptimal machine code for each platform. super cute! exactly how i'd have built it. fun to hack on. probably also full of rope-y exploit…

triangle company created a new "system" programming language "for agents" called zero. i love me some new PLs. it's very cute. looks like the mach-o emitter is broken tho. at least from source. https://zerolang.ai/

Published: May 16, 2026 13:19

triangle company created a new "system" programming language "for agents" called zero. i love me some new PLs. it's very cute. looks like the mach-o emitter is broken tho. at least from source. zerolang.ai/

RT by @badlogicgames: I was just going to go fix clipboard stuff in https://github.com/badlogic/pi-diff-review/ via https://github.com/hazat/glimpse and @badlogicgames and @DanielGri did it a few hours ago! Thanks guys! Incorporated into my rip-off that runs on codex and claude (as well as pi) https://github.com/dbachelder/slop-review

Published: May 16, 2026 05:12

I was just going to go fix clipboard stuff in github.com/badlogic/pi-diff-… via github.com/hazat/glimpse and @badlogicgames and @DanielGri did it a few hours ago! Thanks guys! Incorporated into my rip-off that runs on codex and claude (as well as pi)…

R to @badlogicgames: not that it matters, as you all are using gerpertee and clawd as your personal life coach now anyways. might as well go to one of the booths with sama's eye scanning devices and get it over with.

Published: May 15, 2026 22:57

not that it matters, as you all are using gerpertee and clawd as your personal life coach now anyways. might as well go to one of the booths with sama's eye scanning devices and get it over with.

jebus don't do this. why would you give them that data? holy cow.

Published: May 15, 2026 21:10

jebus don't do this. why would you give them that data? holy cow. ChatGPT (@ChatGPTapp) A preview for Pro users: a new personal finance experience in ChatGPT. Pro users in the U.S. can securely connect financial accounts, see where their money is…

RT by @badlogicgames: Coming off of this psychosis and it’s painful. Not shipping any more features until I can explain every part our codebase in detail. Especially important to do before we open source

Published: May 15, 2026 20:27

Coming off of this psychosis and it’s painful. Not shipping any more features until I can explain every part our codebase in detail. Especially important to do before we open source Mitchell Hashimoto (@mitchellh) I strongly believe there are entire…

RT by @badlogicgames: This is now pushed on the repository. The 75 tests are 25 each for GPQA Diamond / Super GPQA / AIME2025. They are roughly ordered by complexity. In the picture: 4 bit. 2 bit scores a bit worse, not dramatically \o/. This test is not designed to be all-passed, it puts the model at play with hard Q. Useful to improve GGUF files and catch inference errors ASAP.

Published: May 15, 2026 18:45

This is now pushed on the repository. The 75 tests are 25 each for GPQA Diamond / Super GPQA / AIME2025. They are roughly ordered by complexity. In the picture: 4 bit. 2 bit scores a bit worse, not dramatically \o/. This test is not designed to be…

sneak peak of the new http://pi.dev no tools mode. you're gonna love it. you can try it with `pi -nbt`. enjoy!

Published: May 15, 2026 17:53

sneak peak of the new pi.dev no tools mode. you're gonna love it. you can try it with `pi -nbt`. enjoy! Video Mario Zechner (@badlogicgames) people of pi.dev. i'm removing all tools from pi witbout replacement. get creative. —…

> The reason is simple: everybody is being inundated by the slop machine. the future is glorious. the future also sucks. https://turso.tech/blog/the-wonders-of-ai

Published: May 15, 2026 17:44

> The reason is simple: everybody is being inundated by the slop machine. the future is glorious. the future also sucks. turso.tech/blog/the-wonders-…

~ 92 additional posts are not shown ~