Dillon Mulroy
@dillon_mulroy
Followers
16K
Following
100K
Media
2K
Statuses
17K
principal engineer @cloudflare • tweets about functional programming, typescript, neovim, & formula 1 • https://t.co/6Vr6DiQmze • https://t.co/4c0WNBRDIp
Raleigh, NC
Joined October 2011
life update... very excited to share i'll be joining @cloudflare to work on the developer platform and workers! the momentum and incredible work that the team has shipped over the past few months has been inspiring and exciting to watch. stoked to get to be a part of it
184
25
1K
every once in a while there's a post saying something like "GUIs are better, people only like TUIs because they want to feel cool" then i tell them to do opencode --port=4096 and go to localhost:4096 they can't fathom that you can build both because they can't do either
1
2
49
metaprogramming is so nice. being able to compile out entire sub systems of code based on a feature flag is dope. in jj-startship i have a feature-flag to optionally compile out git support if you are exclusively using jj
0
0
15
gm. lets sling some code today
2
0
14
@samwhoo i see all of you bookmarking this with the intent to read it later; don't. open it now, take 10 minutes of dedicated time and focus to read it. you'll be glad you did.
1
0
17
while it teaches the science behind prompt caching, @samwhoo's blog post is, imo, the best intro to how LLM's work. filled with Sam's charm, animations, and ability to distill highly technical topics into engaging blog posts, this is well worth your time https://t.co/SyC4ysYRgC
ngrok.com
A far more detailed explanation of prompt caching than anyone asked for.
5
7
121
just published jj-starship a simple tool optimized to display jj or git info for https://t.co/EHDRsadYkZ cargo install jj-starship https://t.co/iQ0Luhzjcb
10
3
111
i need to process 10,000 files with bounded concurrency, progress tracking, error isolation, backpressure, and streaming results effect is cheating
23
6
164
mark your calendars! excited to go deep on how Dimitri built doom with just typescript types. reply with any questions you want me to ask
got Doom running on an @Ableton Push! prepping for the full technical deep dive on the Doom-in-TypeScript-Types codebase with @dillon_mulroy on Jan 4th at 1pm Eastern (FINALLY🥹)!
1
1
49
In the latest version of Oxfmt (better prettier replacement): • oxfmt --migrate prettier • oxfmit --init • oxfmt --stdin-filepath for pipe usage • Node.js API • Sort package.json fields by default • Respect root .editorconfig Give it a try: •
13
25
389
Been working on a different type of project, I built this with and for my amazing wife. @dillon_mulroy peer pressured me into showing it yesterday😂 but it is ready for an early look, let me know what you think and share it with the normies in your life https://t.co/gVXdooUF2z
0
2
12
i am convinced that software devs have a speed problem they think the #1 issues is writing code faster... its not. its fixing the code that is already there to stop being utter garbage (as a garbage code connoisseur) quality is really lacking these days, yet quantity has never
303
308
5K
Static prerendering OOTB with Start on workers now :) https://t.co/s3hPISlRBl
developers.cloudflare.com
TanStack Start apps can now prerender routes to static HTML at build time using the Cloudflare Vite plugin.
11
16
249
Do things quickly --- That thing that you're convinced will take 6 months? DOA. Instead, make a plan that ships most of it in 6 weeks. You'll get the buy-in for 6 months if it's worth it. 6 weeks to delivery? Break it down into 6 1-week sprints. Setup a quick review every
15
36
530
firefox watched the browser company fumble on arc and said hold my beer
Something that hasn't been made clear: Firefox will have an option to completely disable all AI features. We've been calling it the AI kill switch internally. I'm sure it'll ship with a less murderous name, but that's how seriously and absolutely we're taking this. …
2
0
49
duuuuuurable objects
We’re hitting a massive wall in the AI stack: serverless backends are choking on agents. Developers are hitting execution ceilings and resorting to network hacks just to keep their apps alive and responsive. Reliable agentic loops require persistent, long-running infrastructure
6
8
131