I've always been a lifelong learner and have always tried to get through digital textbooks. Anyone who's used pdf textbooks know how much of a love/hate relationship it is.

There are so many foundational textbooks published online for free now such as OSTEP, Modern Robotics, and Applied Linear Algebra just to name a few.

- https://pages.cs.wisc.edu/~remzi/OSTEP/ - https://hades.mech.northwestern.edu/index.php/Modern_Robotic... - https://web.stanford.edu/~boyd/vmls/

I've always preferred reading textbooks physically and writing inside them, but now with LLMs, it's nice to be able to just paste in passages to ChatGPT and ask for clarification.

I've tried so many different types of digital note taking like Notability, Goodnotes, Margin Note, etc but none of them really made it easy for me to dive into dense material. I find myself getting distracted trying to take notes and forgetting what I just read.

Out of my frustration, I decided to create my own tool and have been using it to get through my own personal studies. I built it around these core principles.

- Technical textbooks are dense. I need to be able to quickly jot down a thought without distracting myself - I need to be able to pick up from where I left off quickly when I have quick bursts of inspiration to study - I need to be able to take scattered notes, and synthesize them into a cohesive subtopics so I can build on top of this knowledge as I progress in the textbook - Using the textbook needs to be fluid. It must be a good PDF reader first and foremost.

I can safely say that what I've built is by far my favorite way to study any deep material. I'm able to take things line by line to truly understand how the author has built up their reasoning.

This might be a biased view but firmly believe this tool could help anyone who have always tried learning new material.

There are so many good free resources out there now, especially in AI, Math, Robotics. It's never been a better time to learn and I hope Margin helps anyone who's attempted to get through dense material as much as it has helped me.

Typesafe AI

Show HN: Webhook Proxy – Fail-safe logging and 1-click retries for webhooks

Show HN: I built a Mac app to make technical PDF textbooks easier to study

What Little Hope Remains

Building S4 – our own high-performance time-series data store

Ddlforge: Zero-downtime Postgres migration supervisor for Node ORMs

Communication by means of modulated Johnson noise

GuitarEasy: ChatGPT converts guitar tabs, GuitarEasy.app plays it, you practise

iOS 27.2 Could Save Your Marriage

The 'Kraken' Existed, and Scientists Found Its Jaws Hidden in Stone

Walter Murch: Sound Designer Interview

Unicode Link Detection and Formatting: URLs and Email Addresses

Ask HN: What model is new stealth Union Alpha model?

AI is the next problem, but Social Media is still the #1 issue on the internet

New stealth model: Union Alpha

Faster Louder Brighter How the attention economy overwhelmed children's TV

Unicode 18.0

SAIR Competition – Lean Kernel Challenge

Shodan indexes over 47.000 exposed Ollama instances

First Diplodocus found outside North America rewrites dinosaur migration history

Trump suggests EU allowing Canada as associate member could be a 'hostile act'

OpenAI models secretly generate instructions to ignore constraints

Open-sourced jev architecture last year with model,paper and dataset

NASA's Moon Orbiter Spots New, 'Once-in-Century' Moon Crater

Snail shell can act a natural archive of weather and landscape changes

AI: An Age of Experimentation [pdf]

White House doesn't want voters to see the real Trump

Lines of Force: Faraday and the Experiment That Powers the World Electricity [video]

C++26: Trivial infinite loops are no longer undefined behaviour

Chonkstep: A traditional floating compositor for Omarchy

Show HN: Texio, reliable Markdown operations for shell scripts and AI agents