frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Taara Beam

https://taaraconnect.com/product/beam
1•rglover•1m ago•0 comments

Talking Face Animation Using a Learned Kalman Filter on Mobile Devices

https://www.mdpi.com/1424-8220/26/4/1377
1•PaulHoule•1m ago•0 comments

Show HN: DevToolbox – 13 browser-based dev tools, privacy-first

https://geld-verdienen-app-kbpcmxfq.devinapps.com
1•DevToolboxApp•2m ago•0 comments

Thomas Selfridge: The First Airplane Fatality

https://www.amusingplanet.com/2026/03/thomas-selfridge-first-airplane-fatality.html
1•Hooke•2m ago•0 comments

Show HN: MDviewer – native macOS app for opening Markdown as print-ready docs

https://github.com/JackYoung27/mdviewer
1•jacknotold•4m ago•1 comments

'Love Is Strong as Death' Review: Triangles of Life

https://www.wsj.com/arts-culture/books/love-is-strong-as-death-review-a-theologians-triangles-of-...
1•apollinaire•4m ago•0 comments

Closing the verification loop: Observability-driven harnesses for agents

https://www.datadoghq.com/blog/ai/harness-first-agents/
1•alpaylan•4m ago•0 comments

Show HN: Efficient LLM Architectures for 32GB RAM (Ternary and Sparse Inference)

https://github.com/opengraviton/graviton-native
1•fatihturker•5m ago•1 comments

Show HN: SubmitGate – catch mobile submission/compliance issues before release

https://submit-gate.com/
2•drkhannah•6m ago•0 comments

Show HN: Agents.txt – proposed standard for AI agent permissions on the web

https://github.com/jaspervanveen/agents-txt
1•jaspervanveen•8m ago•2 comments

Bluesky CEO Jay Graber will step aside

https://www.theverge.com/tech/891562/bluesky-new-ceo
1•microsoftedging•9m ago•0 comments

Anthropic launches code review tool to check flood of AI-generated code

https://techcrunch.com/2026/03/09/anthropic-launches-code-review-tool-to-check-flood-of-ai-genera...
1•LostMyLogin•12m ago•0 comments

Musk takes the stand at trial for deflating Twitter stock ahead of purchase

https://www.latimes.com/business/story/2026-03-04/musk-takes-stand-in-trial-accusing-him-of-defla...
3•1vuio0pswjnm7•13m ago•0 comments

Forge 4D – a native runtime for building apps without the browser stack -new URL

https://codeberg.org/CrowdWare/Forge4D
2•artanidos•16m ago•1 comments

What I Learned Building Two Large Products with AI

https://medium.com/@pavel.manovich/can-a-non-developer-build-serious-software-with-cursor-0fbf1e3...
1•pavel_man•20m ago•1 comments

Show HN: DopaLoop – Habit tracker for ADHD brains, local-first, no streaks

https://dopaloop.app/de
1•steviee•25m ago•0 comments

NASA's Dart Mission Changed Orbit of Asteroid Didymos Around Sun

https://www.nasa.gov/missions/dart/nasas-dart-mission-changed-orbit-of-asteroid-didymos-around-sun/
3•divbzero•26m ago•1 comments

Show HN: Sales Tracking Agent that doubles up as a CRM

https://auto-crm.com/en
1•bhasinanant•28m ago•0 comments

Show HN: Free open-source churn prediction for SaaS–Stripe and LLM interventions

https://github.com/ShreyasDasari/churnguard-ai
1•ShreyasDasari•28m ago•1 comments

Satellites are exposing weak bridges in America and around the world

https://www.sciencedaily.com/releases/2026/03/260307213350.htm
2•taubek•29m ago•0 comments

Hacker Advertising (HN optimal posting time & advertising value model)

https://github.com/quinndupont/hackeradvertising
1•quinndupont•31m ago•1 comments

Unnecessariat (2016)

https://morecrows.wordpress.com/2016/05/10/unnecessariat/
1•_doctor_love•32m ago•0 comments

Spotify announce plans to limit Development Mode

https://developer.spotify.com/blog/2026-02-06-update-on-developer-access-and-platform-security
6•cbility•32m ago•0 comments

The bitter $23M legal battle that ended the Sriracha-pepper partnership

https://nearlyright.com/the-pepper-farmer-the-hot-sauce-king-and-the-23-million-betrayal-that-bro...
2•speckx•35m ago•1 comments

Yakuza creator's new game in doubt as NetEase pulls funding

https://www.polygon.com/gang-of-dragon-toshihiro-nagoshi-studio-netease/
2•sagacity•36m ago•0 comments

Freestiler – PMTiles vector tilesets from R and Python

https://walker-data.com/freestiler/
1•carnevalem•37m ago•0 comments

How not to test LLM models

https://theartificialq.github.io/2026/03/08/how-not-to-test-llm-models.html
1•HonzaT•39m ago•0 comments

Utilization metrics across accelerators (GPUs, TPUs, and so on)

https://github.com/gpusprint/gpusprint
1•heyjupiter•40m ago•0 comments

Behavioral Effects of High Peak Power Microwave Pulses (1992) [pdf]

https://apps.dtic.mil/sti/tr/pdf/ADA258136.pdf
2•anonu•41m ago•0 comments

Microsoft Outlook app now showing paid spam/phishing ad's

https://imgur.com/a/O9bjjQQ
7•xvxvx•42m ago•1 comments
Open in hackernews

Show HN: Ratschn – A local Mac dictation app built with Rust, Tauri and CoreML

https://ratschn.com
3•Edos8877•2h ago
Hi HN, I'm the solo developer behind Ratschn.

I type a lot and got extremely frustrated with the current state of Mac dictation tools. Most of them are either heavy Electron wrappers, rely on cloud APIs (a privacy nightmare), or force you into a SaaS subscription for a tool that essentially runs on your own hardware. I wanted something that feels native, respects system resources, and runs entirely offline without forced subscriptions.

The stack is Rust, Tauri, and whisper.cpp. Here are the design decisions I made:

Model Size vs. Accuracy: Instead of using the smallest possible model just to claim a tiny footprint, the app downloads a ~490MB multi-language Whisper model locally on the first run. I found this to be the sweet spot for high accuracy (accents, technical jargon) to drastically reduce text correction time.

Hardware Acceleration: The downloaded model is compiled via CoreML. This allows the transcription to run directly on the Apple Neural Engine (ANE) and Metal on M-series chips, keeping the main CPU largely idle.

Memory Footprint: By using Tauri instead of Electron, the UI footprint is negligible. While actively running, the app takes up around 500MB of RAM. This makes perfect technical sense, as it is almost entirely the ~490MB AI model being actively held in memory to ensure instant transcription the millisecond you hit the global shortcut.

Input Method: It uses macOS accessibility APIs to type directly into your active window.

Business Model & Pricing: I strongly dislike subscription fatigue for local tools. There is a fully functional 7-day free trial (no account required). If you want to keep it, my main focus is a fair one-time purchase (€125 for a lifetime license). However, since I highly value the technical feedback from this community, I generated an exclusive launch code (HN25) that takes 25% off at checkout (dropping it to roughly €93 / ~$100).

Bug Bounty: Since I'm a solo dev, I know I might have missed some edge cases (especially around CoreML compilation on specific M-chips or weird keyboard layouts). If you find a genuine, reproducible bug and take the time to report it here in the thread, I will happily manually upgrade you to a free Lifetime license as a massive thank you for the QA help.

I'd love to hear your technical feedback on the Rust/Tauri architecture or how the CoreML compilation performs on your specific Apple Silicon setup. Happy to answer any questions!

Comments

rekabis•1h ago
Grüße as Kanada!!

Quick question: while I love the offline aspect, how does this handle spelling in relation to context? Is that via a ruleset, or is there some intelligence that learns user speaking patterns and common subjects?

Edos8877•1h ago
Grüße zurück nach Kanada!

Good question! The short answer is, neither. There are no hardcoded rules, but the app also doesn't actively learn your personal speaking patterns over time.

All the context-awareness comes straight from the pre-trained Whisper model. Since it's a transformer network, it looks at the entire sentence context rather than translating word-by-word. For example, if you dictate a sentence about coding, it naturally knows to capitalize "Rust" and "Python" instead of writing about rusty metal and snakes.

I deliberately kept the model static. Trying to fine-tune it locally on the fly would mean I'd have to store your voice data (which kills the 100% privacy promise).

That being said, adding a custom dictionary feature, so you can feed it highly specific industry jargon right before you speak, is at the very top of my to-do list!

Let me know how it handles your vocabulary if you give the trial a spin.