frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

How the 'math wars' in Palo Alto reflect a raging nationwide debate

https://www.sfchronicle.com/bayarea/article/math-class-palo-alto-school-district-22478798.php
1•randycupertino•5m ago•1 comments

LiDAR Lens: web viewer for 1M Lidar terrain across England and Wales

https://lidarlens.sarsen.org
1•winterswift•7m ago•0 comments

Chinese research vessels loiter near undersea cables more than any other nation

https://www.tomshardware.com/tech-industry/chinese-research-vessels-loiter-near-pacific-undersea-...
2•beansovertoast•10m ago•0 comments

The 'Buy, Borrow and Die' Wealth Strategy Has Become a $5T Market

https://www.wsj.com/finance/the-buy-borrow-and-die-wealth-strategy-has-become-a-5-trillion-market...
1•tysone•15m ago•0 comments

Accessory dwelling unit ruling could be a boon to California builders

https://www.sfchronicle.com/politics/article/adu-ruling-boon-california-builders-22495410.php
1•littlexsparkee•18m ago•0 comments

Show HN: An MCP server that writes and sings songs in 76 languages

https://www.songupai.com/mcp-server
1•rakim777•20m ago•0 comments

How Australia Built World’s Biggest Battery-Powered Ferry

https://www.bloomberg.com/news/features/2026-10-26/an-83-year-old-shipbuilder-is-delivering-era-o...
1•toomuchtodo•23m ago•1 comments

OpenXR Spatial Container Extensions Standardize Volumetric App Multitasking

https://www.uploadvr.com/openxr-spatial-container-extensions-standardize-volumetric-app-multitask...
1•andsoitis•24m ago•0 comments

What to do if you discover AI models trained using your copywrighted book?

4•andrea81•26m ago•3 comments

These Neurons Control Territorial Claims

https://edwardbenson.com/tinkering/neuron-explorer
1•eob•27m ago•0 comments

GLM-5.3 Sometimes Considers How Train-Time Actions Shape Future Behavior

https://www.lesswrong.com/posts/R5g8YT25G9DXJJ48n/glm-5-3-sometimes-considers-how-train-time-acti...
2•eatitraw•28m ago•0 comments

16-bit Windows (Windows 1.x, 2.x, 3.0, 3.1, etc.) on 64-bit Windows

https://github.com/otya128/winevdm
7•doener•30m ago•1 comments

Do you want to keep your job? Stop feeding AI models

2•andrea81•31m ago•3 comments

Initial Support for the Apple M4 and Apple A18 Pro Make It into Linux 7.4

https://www.phoronix.com/news/Linux-7.4-SoCs
4•doener•32m ago•0 comments

Show HN: Libtmux-ts 0.1 – tmux library for Node.js, TypeScript, bun

https://github.com/libtmux/libtmux-ts
1•tony•35m ago•0 comments

AMA – "Final Fantasy: The Spirits Within", milestone in CGI, is 25 years old

https://old.reddit.com/r/FinalFantasy/comments/1xgd2qy/ama_final_fantasy_the_spirits_within_miles...
2•kaveh808•35m ago•0 comments

Engineering Explained: I Fact-Checked Every Stunt in the Tesla Roadster Reveal [video]

https://www.youtube.com/watch?v=hJSbro3EBnY
2•lathiat•37m ago•3 comments

Let your agent talk to mine

https://blog.danielbrain.ai/let-your-agent-talk-to-mine
1•bluepnume•37m ago•0 comments

AI wasn't overhyped. Almost everyone underestimated it

https://benjamintodd.substack.com/p/almost-every-ai-forecast-of-the-last
2•gmays•38m ago•0 comments

Show HN: Sparta – A $799 starting point before Vanta

https://marathon.sh/sparta
1•shotgunsparkle•41m ago•1 comments

Show HN: Torsher – a self-hosted, TV-first media center in one container

https://github.com/torsher-app/torsher
1•antonreshetov•44m ago•0 comments

The "let-it-rip" camp is wrong on AI

https://economist.com/by-invitation/2026/10/26/the-let-it-rip-camp-is-wrong-on-ai
1•andsoitis•44m ago•1 comments

Show HN: Marlow – Build full-stack apps with AI and own your code

https://marlow.app/beta
1•kristapsmarlow•45m ago•0 comments

Request for Immediate Release of Yang–Mills AI Output

https://agmai.org/#request-immediate-release
1•kerim-ca•54m ago•0 comments

From RTL and No-RTL to GDS in 3 Weeks

https://www.ustcpetergu.com/MyBlog/experience/2026/10/10/from-rtl-and-no-rtl-to-gds.html
1•paaloeye•55m ago•1 comments

Using Env(1) in the Shebang

https://notes.pault.ag/env-in-shebang/
3•ykgoon•56m ago•0 comments

Why using AI saves doctors time but won't fix physician burnout

https://theconversation.com/why-using-ai-saves-doctors-time-but-wont-fix-physician-burnout-291250
1•beansovertoast•56m ago•0 comments

Roblox downloaded 145 GB of models from our sites without asking

https://open3dlab.com/news/roblox-downloaded-145-gb-of-models-from-our-sites/
2•linksbro•57m ago•0 comments

AI Crisis in Math: Navier-Stokes, Riemann Hypothesis and Complexity Theory [video]

https://www.youtube.com/watch?v=ZAzWPE9CItg
2•vismit2000•58m ago•1 comments

CPU Scheduling Explained [video]

https://www.youtube.com/watch?v=gq6g31usk0k
1•vismit2000•58m ago•0 comments
Open in hackernews

Show HN: A native Hacker News reader with integrated todo/done tracking

https://github.com/haojiang99/hacker_news_reader
54•coolwulf•1y ago
Hey HN! I'm excited to share a tool I've been working on - a native Hacker News reader built with Rust and egui. Here's a screenshot: https://github.com/haojiang99/hacker_news_reader/blob/main/s....

As a daily HN reader, I've always struggled with keeping track of interesting posts I want to read later. Browser tabs pile up, bookmarks get forgotten, and I lose track of what I've already read. I needed a way to:

1. Browse HN efficiently (across all sections - hot, new, show, ask, jobs, best) 2. Quickly mark posts as "todo" for later reading 3. Mark posts as "done" when finished 4. Filter and search effectively

I couldn't find a tool that combined all these features, so I built one. It's been tremendously helpful for my own HN reading workflow, and I thought others might find it useful too.

Features:

- *Integrated todo tracking*: Mark stories as "todo" and "done" to manage your reading progress

- *Search functionality*: Filter stories by keyword in title, domain, or author

- *Multiple sections*: Browse all HN sections (hot, new, show, ask, jobs, best)

- *Threaded comments*: View comments in a Reddit-like threaded format

- *Dark/light mode*: Easy on the eyes in any environment

- *Keyboard shortcuts*: Efficient navigation with keyboard-centric design (1-6 for tabs, Ctrl+F for search)

- *Auto-loading*: Automatically loads more content when scrolling

- *Color-coding*: Stories color-coded by score for easy scanning

- *Native app*: Fast, responsive, and works offline with local caching

Built with Rust and the egui UI framework, with SQLite for local storage. The app scrapes Hacker News HTML directly rather than using the official API to capture the full story context.

Check out the GitHub repo (https://github.com/yourusername/hacker_news_reader) for installation instructions and source code. Built and tested on macOS, Linux, and Windows.

This started as a personal tool to solve my own HN reading habits, but I hope others find it useful too. The code is MIT licensed and I'd love your feedback, feature suggestions, or contributions!

Comments

Beijinger•1y ago
OK, why not RSS reader?
coolwulf•1y ago
Because I can customize a lot of functions related to Hacker News which usual RSS reader cannot
Beijinger•1y ago
Yes, this is obvious. But why don't you offer YOUR script not for a general RSS feed?
coolwulf•1y ago
That could be another tool :p
shahakshat609•1y ago
sounds like a good idea & a lot of people who would need this
coolwulf•1y ago
Thank you. I will work to make it available to more people and make this system more powerful and extremely fast.
dsr_•1y ago
Every forum system tries to replicate the user-empowering features of Usenet with slrn, and usually falls short.
coolwulf•1y ago
Well I'm implementing a LLM based auto summary for each article and other features which focus on usability / readability / tracking for HN users. And this is more specialized and cannot be achieved by other tools.
dsr_•1y ago
Hrm. You should pivot to making an Electric Monk.

"The Electric Monk was a labour-saving device, like a dishwasher or a video recorder. Dishwashers washed tedious dishes for you, thus saving you the bother of washing them yourself, video recorders watched tedious television for you, thus saving you the bother of looking at it yourself; Electric Monks believed things for you, thus saving you what was becoming an increasingly onerous task, that of believing all the things the world expected you to believe."

- Douglas Adams.

flave•1y ago
Perhaps you could teach the LLM to perform the task of enjoying the articles for you too?
barbazoo•1y ago
> https://github.com/yourusername/hacker_news_reader

Link is dead

coolwulf•1y ago
Please use the thread title link
pricci•1y ago
AI artifact?

The repo is https://github.com/haojiang99/hacker_news_reader

barbazoo•1y ago
What I thought yeah. It’s LLM generated shit all the down.
beemboy•1y ago
Building an HN reader feels like the Hello World for hackers. Still waiting for HN in BF running in the Doom HUD.
coolwulf•1y ago
Pretty much. That is why this was a tiny personal project I did. The key for me is to be a le to track article and comments on HN for myself. I found it useful and that's why I decided to open source and make it available for everyone else.
krishnasangeeth•1y ago
Cool project. I too did something similar , built a HN client app for iPhone

https://apps.apple.com/us/app/h-for-hackernews/id6744145932

begueradj•1y ago
The second Github link is broken