Hawk Ticehurst

@hawkticehurst.com

UX Engineer @vscode.dev. He/Him 🕸️ hawkticehurst.com 📍 New York, NY

💥 just released a github action: `uppt` it aims to be a very simple, secure release workflow for maintainers that adheres to best security practices ⚪ trusted, staged publishing on npmjs.com ⚪ github environment protection ⚪ generated release PR + changelog ⚪ automatic release + publish on merge ✨

Mythos' underwhelming results for curl are a great example of why Open Source is even more important now that we have access to these tools. curl is more secure because everyone can tinker with it. We should be doubling down on Open Source rather than closing projects in panic.

daniel:// stenberg://@bagder.mastodon.social.ap.brid.gy · 3mo ago

#Mythos finds a #curl vulnerability yes, as in singular one. https://daniel.haxx.se/blog/2026/05/11/mythos-finds-a-curl-vulnerability/

A year of work later and we can finally tie a bow on this series of web component perf threads. I built a fully declarative, compile-time reactive framework on web components. In the latest JS Framework Benchmark, Roqa is the 2nd fastest (unflagged) – beating Ripple, Vue Vapor, Solid & Svelte. 🏎️💨

Screenshot of the official JS Framework Benchmark run for Chrome 148. The implementation for Roqa is highlighted, showing it is in the second place position.
Hawk Ticehurst@hawkticehurst.com · last yr.

The idea of web component “overhead” is overblown — to even my own surprise! Rendering 1000s of nested WCs is *more* than competitive with the heavyweight champs of the web framework world. We’re talking S/S+ tier perf depending on the initialization method. Cont…

Screenshot of the JS Framework Benchmark results run locally on my computer. The order of results from most performant to least is:

1. Vanilla JS
2. Web Components 1
3. Web Components 2 Fast
4. Web Components 2
3. Web Components 3 Fast
4. Svelte 1
5. Solid 2
6. Solid 1
7. Web Components 3
8. Solid 3
9. Svelte 2
10. Svelte 3

Last year I prototyped a feature in VS Code where your entire chat thread was editable, meaning you could edit mistakes, cut out bad context, etc. It was fun to play with but from my experience it was kind of a wash for improving performance. Most of the time starting a fresh chat was better.

David Buchanan@retr0.id · 3mo ago

I wonder if you could improve agent performance by editing out corrected mistakes from context retroactively (as if it always got things right first try)

🌶️: If we were still writing all our apps in vanilla web tech in 2026, the state of the web would be at least 100x worse than what it is today. React's model won because good abstractions are good! Creating better abstractions is like kind of the *entire* deal when it comes to programming / software

dan@danabra.mov · 4mo ago

it’s always funny to me when people try to paint react as some kind of “big company” propaganda. it was like one person’s wild idea and then there was at most ~5 people maintaining it for years. people HATED it when fb open sourced it. the only reason it won is because the model was actually good

NASA and Mamdani are proving that Americans are desperate to see their money spent in ways they feel good about. This, however, is not the same thing as simply spending the money on things we would feel good about but never know happened. People need to be able to see it.

On my 2030s bingo card: ATProto eventually becomes *the* desirable and premium app ecosystem because if you enshittify your app you die The apps that win will win because they’re literally the best or most specialized at what they do

Guido X Jansen@gui.do · 4mo ago

That means we need new moats. Deeper specialization. Better UX. Domain expertise. Community trust. Optional premium features. Enterprise integrations. The old thinking of "my proprietary access to data protects me" is exactly the Big Tech model we're here to leave behind.

Y'all are sleeping on VSCode's AI UX. Their velocity here is wild, they keep building so many great features that improves the agent experience. I also think GitHub Copilot is fantastic, but if Claude Code is more your jam, you can use custom agents now within VSCode's chat, which is pretty cool.

Visual Studio Code@vscode.dev · 6mo ago

You told us you’re running multiple AI agents and wanted a better UX. We listened and shipped it! Here’s what’s new: 🗂️ Unified agent sessions workspace for local, background, and cloud agents 🔀 Parallel subagents 💻 Claude and Codex support for local and cloud agents 🌐 Integrated browser & more...

Got to watch IRL and it's easily one of my favorite talks / takes on AI-driven development I've seen so far Rich (as always) wonderfully threads the needle of taking a critical look at this topic without being AI-pilled or anti-AI. He's nuanced, funny, and treats the topic with respect

rich harris@rich-harris.dev · 6mo ago

I cover quite a lot of ground in this talk, but the bit I hope people take away from it is that AI-driven development poses dangers that we're not talking about nearly enough. This clip was recorded before we knew how prominently Elon features in the Epstein files: 🎞️ youtube.com/clip/UgkxImh...

You told us you’re running multiple AI agents and wanted a better UX. We listened and shipped it! Here’s what’s new: 🗂️ Unified agent sessions workspace for local, background, and cloud agents 🔀 Parallel subagents 💻 Claude and Codex support for local and cloud agents 🌐 Integrated browser & more...

Depending on how the winds of the future blow, a social filesystem could also be the foundation for an entirely new computing platform where your OS is portable / device agnostic Like @rsms.me’s Playbit, but anyone could come along make a Playbit competitor and still access the same filesystem/data

𝕮@chrisshank.com · 7mo ago

The foundation of personal computing is the filesystem. The foundation of the internet is a networked filesystem. The foundation of a new social web will likely be a social filesystem. (It remains an ope question if atproto is such a foundation)