Hacker News

Article URL: https://www.dataandpolitics.net/trump-is-a-critical-vulnerability/

Comments URL: https://news.ycombinator.com/item?id=43726065

Points: 2

# Comments: 1

Article URL: https://www.nature.com/articles/d41586-025-01129-5

Comments URL: https://news.ycombinator.com/item?id=43726063

Points: 1

# Comments: 0

In February I posted a rough AI browser‑game prototype and asking game dev tips. Then I got about 100 comments that really helped. I went back, rewrote most of it, and here’s the current state:

What’s new - Endless world with mountains, lakes, and altitude‑based snow - Living systems – trees grow, bears roam and charge if you trespass - View switch – press V for first‑ or third‑person on the fly - Guest mode – no login, just play - Type‑anything crafting – enter “wooden glider,” “big wall,” “steak,” etc with proper materials; AI invents the recipe and stats

Still many things are raw and not-ready, but I want to keep sharing progress and gathering feedbacks.

If you try it, I’d love to hear where it falls apart first and whether the core loop feels fun.

Thanks for any feedback!

Comments URL: https://news.ycombinator.com/item?id=43726058

Points: 3

# Comments: 0

Article URL: https://defold.com

Comments URL: https://news.ycombinator.com/item?id=43726051

Points: 1

# Comments: 0

Hi HN!

I've noticed that as the applications of LLMs continue to expand, the demand for tokens is increasing, which also brings significant cost burdens to users. Currently, there aren't many effective prompt compression products on the market, so I've created a simple demo site: https://hello-llm.com.

This application includes two features: lossy compression and lossless compression.

1. Lossy compression allows users to specify the compression ratio and is faster; 2. Lossless compression lets the model determine the compression ratio, aiming to retain as much useful information as possible, albeit slower.

This is just a demo, and I plan to assess its feasibility and gather enough feedback before deciding whether to launch a formal product. Everyone is welcome to try it out!

Comments URL: https://news.ycombinator.com/item?id=43726039

Points: 2

# Comments: 0

Article URL: https://dl.acm.org/doi/10.1145/3552326.3587458

Comments URL: https://news.ycombinator.com/item?id=43726037

Points: 1

# Comments: 0

Article URL: https://www.mycarousel.fun/

Comments URL: https://news.ycombinator.com/item?id=43726031

Points: 2

# Comments: 0

Article URL: https://arxiv.org/abs/2504.05411

Comments URL: https://news.ycombinator.com/item?id=43726013

Points: 1

# Comments: 0

Article URL: https://github.com/hxu296/tariff

Comments URL: https://news.ycombinator.com/item?id=43726012

Points: 2

# Comments: 0

Article URL: https://www.militarystory.org/how-the-military-hid-the-lockheed-burbank-aircraft-plant/

Comments URL: https://news.ycombinator.com/item?id=43725998

Points: 1

# Comments: 0

The tool helps me keep track of issues with my websites, which I think can be helpful to anyone with a website.

It runs at least once a day and does a lot of checks, such as broken links, bad meta tags, missing H1 and such, but more importantly one of the core focuses is internal linking, which has been shown to be very important to rank on Google.

Always looking for feedback, please let me know!

Here's a limited 80% discount code, first come first served: GIMMESUM

Comments URL: https://news.ycombinator.com/item?id=43725960

Points: 1

# Comments: 0

Article URL: https://arstechnica.com/ai/2025/04/openai-releases-new-simulated-reasoning-models-with-full-tool-access/

Comments URL: https://news.ycombinator.com/item?id=43725917

Points: 1

# Comments: 0

Article URL: https://www.gothpunk.com/articles/open-source-punks.html

Comments URL: https://news.ycombinator.com/item?id=43725913

Points: 1

# Comments: 0

Article URL: https://en.wikipedia.org/wiki/The_Geographical_Pivot_of_History

Comments URL: https://news.ycombinator.com/item?id=43725890

Points: 1

# Comments: 1

Article URL: https://www.thetransmitter.org/learning/dopamine-gas-pedal-and-serotonin-brake-team-up-to-accelerate-learning/

Comments URL: https://news.ycombinator.com/item?id=43725839

Points: 2

# Comments: 0

Article URL: https://hamel.dev/blog/posts/evals/

Comments URL: https://news.ycombinator.com/item?id=43725838

Points: 1

# Comments: 0

Article URL: https://techcrunch.com/2025/04/17/wasps-platform-is-the-glue-that-holds-web-apps-together/

Comments URL: https://news.ycombinator.com/item?id=43725834

Points: 1

# Comments: 0

Article URL: https://www.bugsink.com/blog/why-i-gave-up-on-self-hosted-sentry/

Comments URL: https://news.ycombinator.com/item?id=43725815

Points: 1

# Comments: 0

Hi HN,

I built dir2txt — a simple but powerful CLI tool that turns a directory tree into a clean, structured text or JSON dump.

What It Does • Traverses a project directory • Dumps readable file contents • Optionally strips comments (smart detection of comment blocks + patterns) • Respects .gitignore, .dockerignore, .npmignore, etc. • Outputs LLM-friendly .json or .txt files

Why I Made It

Working with LLMs like GPT-4 or Claude, I kept hitting the same issue: how do I give the model meaningful access to an entire project?

Manually pasting files is noisy and lossy. Existing tools like tree only give structure. I wanted one tool that could: •Structure the directory •Include the real content •Strip boilerplate if needed •Output in a format usable by RAG pipelines, code copilots, or embedders

Example $>dir2txt ./my-project --strip-comments --json > project.json

And then feed it into a vector DB or tokenizer for embedding + retrieval.

Install

Homebrew $>brew tap shubhamoy/dir2txt $>brew install dir2txt

Comments URL: https://news.ycombinator.com/item?id=43725808

Points: 1

# Comments: 0

Article URL: https://citiesandmemory.com/heritage/

Comments URL: https://news.ycombinator.com/item?id=43725800

Points: 1

# Comments: 1