frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Kaliningrad Oblast

https://en.wikipedia.org/wiki/Kaliningrad_Oblast
1•simonebrunozzi•1m ago•0 comments

Probably the least productive container you'll ever run

https://github.com/ahmubashshir/unemployd/
2•bdeshi•6m ago•0 comments

SPA vs. Hypermedia: A Collaborative Kanban Board Measured Against Planka

https://zweiundeins.gmbh/en/blog/spa-vs-hypermedia-a-collaborative-kanban-board-measured-against-...
2•ciconia•6m ago•0 comments

The Problem with C++ Safety: Why "It Works" Isn't Good Enough

https://www.modernescpp.com/index.php/cpp-and-memory-safety-part-1/
2•jandeboevrie•7m ago•0 comments

Show HN: Wordchute – a word game where every word changes the board

https://wordchute.com
2•codercolin•7m ago•0 comments

Side-Effect Decoherence

https://t3.at/blog/2025-08-28-side-effect-decoherence/
1•andsoitis•9m ago•0 comments

Torvalds use Claude for bottleneck fix, but race discovered

https://thenewstack.io/torvalds-claude-code-kernel/
3•PaulDavisThe1st•9m ago•0 comments

Anarchon – Leaderless cron for multi-instance Node.js services

https://github.com/aristoteliss/anarchon
1•aristoteliss•9m ago•0 comments

Replacing an LLM judge with a deterministic model: 9.2x faster, 4.5x cheaper

https://medium.com/@gbenga-awodokun/building-clinical-ai-agents-with-langchain-and-jev-decision-h...
1•newadays•10m ago•0 comments

Trials and Tribulations of Native Lua Compilation

https://samyeyo.github.io/clx/blog.html
1•_samt_•11m ago•1 comments

Show HN: Pi-cowboy – use Codex, Claude Code, OpenCode as subagents

https://github.com/electric-dragon-80000v/pi-cowboy
1•suoloordi•13m ago•0 comments

nmail – Terminal-based email client

https://github.com/d99kris/nmail
1•smartmic•15m ago•0 comments

Show HN: Kramena – Classic books, 20 minutes a day, with spoiler-safe recaps

https://read.kramena.app
1•raahelb•16m ago•1 comments

Building a Compositor for a Userbase of One

https://onewireout.com/writeups/resonance/
1•airhangerf15•19m ago•0 comments

How to Set Up Multiview with RTSP Cameras Using Red5 Pro?

https://www.red5.net/blog/revolutionizing-live-video-streaming-multiview-with-rtsp-cameras-using-...
1•mondainx•19m ago•0 comments

Safety and Liveliness in Software Engineering

https://en.wikipedia.org/wiki/Safety_and_liveness_properties
1•_josh_meyer_•19m ago•1 comments

AI-Assisted Product Discovery: When Code Is Cheap, Judgment Is the Job

https://age-of-product.com/ai-assisted-product-discovery/
1•swolpers•23m ago•1 comments

Python's binary wheels and supply chain attacks

https://rushter.com/blog/python-binary-wheels/
2•birdculture•25m ago•0 comments

Microsoft drops AI datacentre project in Essex over power supply concerns

https://www.theguardian.com/uk-news/2026/oct/18/microsoft-drops-huge-ai-datacentre-project-in-ess...
3•wondersky•30m ago•0 comments

Benchmarking DuckDB v2.0-dev on Nvidia Vera

https://duckdb.org/2026/10/16/nvidia-vera-benchmark
1•tosh•32m ago•0 comments

Ask HN: System One models for autonomous agents

1•manlymuppet•37m ago•0 comments

Campus deplatformings hit record high

https://expression.fire.org/p/campus-deplatformings-hit-record
5•iamnothere•37m ago•0 comments

Simultaneous high-fidelity single-qubit gates in a spin qubit array

https://www.nature.com/articles/s41567-026-03466-2
1•bookofjoe•37m ago•0 comments

Probabilistic Programming

https://adriansampson.net/doc/ppl.html
2•rramadass•47m ago•1 comments

Casio GG-B100X-1A

https://www.casio.com/uk/watches/gshock/product.GG-B100X-1A/
1•asjkff•51m ago•1 comments

Israeli settlers attack European diplomats, beat French mayor

https://www.channelnewsasia.com/world/israeli-settlers-attack-european-diplomats-beat-french-mayo...
9•Markoff•52m ago•1 comments

macOS 15.8.2 Breaks OpenGL

https://developer.apple.com/forums/verify-human/verify.html?next=/forums/thread/850550
49•Audiophilip•52m ago•18 comments

Show HN: AI Zombie Trend–Generating multi-shot narrative videos from two photos

https://aizombietrend.online/
1•dawn3727•53m ago•0 comments

Show HN: Secret Guard – a local LLM gateway; keys never leave your machine

https://github.com/luochen1990/secret-guard
1•luochen1990•53m ago•0 comments

Snow Leopard Was Peak Computer

https://www.alara.blog/snow-leopard-was-peak-computer/
3•asaddhamani•55m ago•0 comments
Open in hackernews

macOS 15.8.2 Breaks OpenGL

https://developer.apple.com/forums/verify-human/verify.html?next=/forums/thread/850550
37•Audiophilip•52m ago

Comments

api•24m ago
These days you could just ask a bot to port your stuff from OpenGL to a newer API.

Backward compatibility is less of an issue when fixing legacy stuff takes $25 in tokens.

kcb•22m ago
Unfortunately there is no universally supported newer API thanks to Apple.
api•21m ago
In that case maybe the answer is $200 in tokens to make a metal OpenGL shim and then share it to the community on GitHub.
mosura•17m ago
Given the comparable performance of OpenGL and Metal on the same hardware I was under the impression that is how Apple had been implementing OpenGL for quite a while.
my123•15m ago
OpenGL on Apple GPUs is implemented on top of Metal with some private extensions to be closer to OpenGL behaviour. Making things work performantly on top of regular Metal is quite a bit harder.
LoganDark•11m ago
Not sure why this is downvoted, it's true: https://rosenzweig.io/blog/asahi-gpu-part-6.html

> Comparing traces, the render pass “OpenGL mode” sets an unknown bit in this kernel-processed block. If we set the same bit in our OpenGL driver, we find the clip space changes to -1/1.

eknkc•11m ago
Isn't https://github.com/google/angle just that?

Edit: I guess no, Angle is OpenGL ES. My bad.

LoganDark•12m ago
Minecraft has been moving to Vulkan and using MoltenVK to run the same backend on Mac.
photios•17m ago
Why, yes, of course, have all the devs in the world spend $25 and tons of time each, because Apple doesn't care about devs nor users.

/s

poszlem•15m ago
How about they spend $25 in tokens to fix what they broke?
drnick1•10m ago
Shouldn't Apple be the one spending the $25 to deliver a product that works?
3form•5m ago
"Why have X when Y costs $N in tokens" is becoming a tired thought-terminating cliche.

It shouldn't be user's responsibility to have to get their binaries to work at all costs.

corysama•21m ago
Seems like a good time to mention that LunarG’s Vulkan-on-Metal wrapper reached Vulkan 1.4 conformance recently.

https://www.lunarg.com/the-state-of-vulkan-on-apple-oct-2026...

Now we need Mesa’s OpenGL-on-Vulkan to be production-ready when running on KosmicKrisp

https://docs.mesa3d.org/drivers/zink.html

mosura•18m ago
The general Apple software quality decline seems to be accelerating. We are in the thousand papercut stage where lots of small things just don’t work properly.

One of my Macs has developed the most insane habit of losing audio output capability, and then tries to route it to another one - a feature I didn’t even know existed.

zitterbewegung•14m ago
I would agree with you if this wasn't a release candidate. I would wait and see at this point.
grumple•7m ago
The response from apple is insane, “please keep trying to reproduce this for us”. That is your job, apple engineer. As a manager of a large software org, a response like that from one of my engineers would drive me insane. Fix the damn problem or provide a timeline for addressing it. When you think the fix is released, then ask them if their issue was resolved.
Razengan•5m ago
You can ask an AI to go through my HN account and see that I used to be an unapologetic Apple fanbitch

Now I must soundly say: Fuck Apple.

Momentum and their hardware quality are the only things setting them apart from everyone, while their ecosystem and individual software products are declining.

There's no way to have ourselves heard by them.

Ironically, our best hope for computing now is that someone eventually uses AI to make a better OS and ecosystem.