frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

CVE-2026-111943: Compromised Intermediate Certificate Authority

https://nvd.nist.gov/vuln/detail/cve-2026-111943
109•keks24•1h ago•29 comments

Faster than earcut: triangulating 3.7M map polygons in 489 ms

https://globus.software/blog/2026/10/16/cutting-map-polygons/
32•molind•1h ago•2 comments

WorldTUI: I trained a 360M model to pretend to be a terminal

https://simedw.com/2026/10/21/world-model-terminal/
232•alphax314•8h ago•70 comments

The recently proposed o(n log n) integer multiplication algorithm [pdf]

https://web.maths.unsw.edu.au/~davidharvey/research/subnlogn.pdf
212•craigcitro•8h ago•144 comments

Security Disclosure

https://blog.railway.com/p/security-disclosure-wednesday-october-21st
474•passenger•12h ago•278 comments

Achieving AI-style reasoning without an LLM

https://fleetingswallow.com/minotaur-ai-deep-dive-achieving-reasoning-without-a-language-model/
30•edwardchalk•1h ago•7 comments

Claude Duet

https://claude.com/product/claude-duet
395•kuberwastaken•15h ago•361 comments

No more icons in every menu by default

https://blog.jim-nielsen.com/2026/icons-in-every-menu/
329•riffraff•2d ago•196 comments

Compromised LiteLLM 1.107.0 on PyPI Installs a Credential Stealer and Backdoor

https://safedep.io/litellm-pypi-compromise-credential-stealer-backdoor/
527•abhisek•20h ago•172 comments

Rust Style Guide

https://google.github.io/styleguide/rustguide.html
316•kaycebasques•15h ago•190 comments

Using LLM for confidence scoring is a bad idea

https://rumple.github.io/blog/2026/10/20/asking-llm-for-confidence.html
101•jimauthors•1d ago•52 comments

GitLab critical patch release

https://docs.gitlab.com/releases/patches/patch-release-gitlab-19-5-1-released/
355•artooro•15h ago•83 comments

OpenAI CLI

https://github.com/openai/openai-cli
128•armcat•5h ago•73 comments

The Robot I Didn't Expect to Build

https://spectrum.ieee.org/agility-robotics-digit-5-redesign
9•bojackh•2d ago•7 comments

What Is the Karakoram Anomaly?

https://nsidc.org/learn/ask-scientist/karakoram-anomaly-it-real
36•thunderbong•2d ago•8 comments

PingCap TiDB Leaks

https://fulcrumsec.vg/pingcap/
352•swyx•1d ago•55 comments

Meta ordered to pay $25B penalty in New Mexico Cambridge Analytica case

https://www.cnbc.com/2026/10/21/meta-new-mexico-cambridge-analytica-penalty.html
572•cbg0•16h ago•330 comments

Tiny possum and glider thought extinct for 6000 years found in remote West Papua

https://www.bbc.com/news/articles/cp814l80pw2o
221•bryanrasmussen•2d ago•51 comments

Code is clay now

https://www.albertodenia.com/writing/code-is-clay-now
179•albertodenia•2d ago•115 comments

mitmproxy adds --emudhra option

https://github.com/mitmproxy/mitmproxy/commit/169234499088fc2c414644a67efaed3d64ae2e3b
488•ahlCVA•10h ago•117 comments

I can't use generic lambdas at compile-time, now what?

https://blog.ganets.ky/ClangGenericLambdas/
14•ibobev•2d ago•1 comments

OpenAI models are coming to Google Cloud

https://openai.com/index/openai-models-google-cloud/
294•armcat•21h ago•121 comments

Startup taps underwater volcano as energy source in "proof point" test

https://arstechnica.com/science/2026/10/startup-taps-underwater-volcano-as-energy-source-in-proof...
5•wffurr•2d ago•0 comments

Beyond the &

https://lwn.net/Articles/1096028/
69•PaulHoule•1d ago•4 comments

DuckDB 2.0.0 LTS “Cinnamon Teal”

https://duckdb.org/2026/10/21/announcing-duckdb-200
422•nnx•20h ago•56 comments

Today's PHP security release, and how we're strengthening security

https://thephp.foundation/blog/2026/10/21/todays-php-security-release-and-how-were-strengthening-...
742•ejohnso49•18h ago•368 comments

ThinkPad I9990303 Error Catch-22

http://www.os2museum.com/wp/thinkpad-i9990303-error-catch-22/
30•userbinator•2d ago•4 comments

Show HN: Omnara – Open-source alternative to Claude Managed Agents

https://github.com/omnara-ai/omnara
82•kmansm27•17h ago•25 comments

Chain of Thought for Humans

https://www.boristhebrave.com/2026/10/18/chain-of-thought-for-humans/
97•ibobev•2d ago•50 comments

There isn't an easy exploit for this PHP vulnerability

https://eliastaylor.com/noexploit/
535•zkldi•11h ago•325 comments
Open in hackernews

A kernel developer plays with Home Assistant

https://lwn.net/SubscriberLink/1017720/7155ecb9602e9ef2/
138•pabs3•1y ago

Comments

balloob•1y ago
Founder Home Assistant here. Want to chime in that I always love to see write ups like these to see the great things what people achieve with Home Assistant.

Not everyone might know, but last year we started the Open Home Foundation[1] as a non-profit in Switzerland and I donated Home Assistant to it[2]. It's fully funded by users. There are no investors involved.

We are fully committed to building out a smart home that focuses on local control and privacy. Yes there are rough edges, but we're actively working on it in the open, with progress being released every month.

~Paulus Founder Home Assistant & President Open Home Foundation https://github.com/balloob

[1]: https://www.openhomefoundation.org [2]: https://www.openhomefoundation.org/blog/announcing-the-open-...

pabs3•1y ago
Discussion for the other article in the series:

https://news.ycombinator.com/item?id=44011381

tomhow•1y ago
Comments moved thither. Thanks!
pabs3•1y ago
They are two different articles, I don't think that was correct.
tomhow•1y ago
The problem is we can’t have two closely-related threads (i.e., threads where there is significant subject/discussion overlap) active at once.

When that happens it just gets confusing, because it’s hard for people know which thread to comment in, if the comment they want to make is somewhere in the overlap. And then whichever one they choose to comment in, people who only see the other thread won’t see that comment. Then sometimes, anticipating this, people will copy and paste their comment in both threads (which happened in this case). But then each one gets different replies.

So each thread ends up being incomplete and duplicated all at once, and it all becomes a big confusing mess.

The fact that these two articles were by the same author, had the same title, were published just a week apart and could easily have been published as one, longer article, says to me that merging the threads was the right thing to do.

The other option would have been to bury the second thread and consider another thread about that second article a few months later, but that didn’t seem like the best option, given how much the two articles are so related and continuous.

Edit: Just thought I'd add that a major factor in deciding to merge the threads was this opening to the second part by the author:

The first article in this series provided an overview of Home Assistant, its community, and its capabilities. It was deliberately short on descriptions of interesting things that can be done with Home Assistant, though — the reasons why one might actually want to use this program. In this closing article, we'll look at how Home Assistant was used to solve some real problems.

To me it makes all the difference that the first part is introductory/high-level whilst the second part goes deeper into usage-scenarios. We'd treat it differently if each part went deeply into different aspects on the project.

pabs3•1y ago
Thanks for the response, guess that makes sense.
pabs3•1y ago
BTW, on lobste.rs, they can merge threads into one, and all the URLs are shown at the top. That might be a useful change to adopt for HN too?