Joined August 2013
30 Photos and videos
Been waiting for this for like 3 years 🥹 Well done
InstancedBatchedSkinnedMesh with GPU mixer. jsfiddle.net/cbenn/Las0poyu
2
1
1
274
Lukewarm take: If you have to keep adding docs for people to understand how to use your APIs, it's probably because your APIs suck Especially if it's because people keep incorrectly assuming what your APIs do, and the assumed behavior is actually reasonable
1
4
157
Derrick Farris retweeted
1 Feb 2024
i miss the old react, straight from facebook react componentDidMount react, with this.state.count react i hate the new react, triangle crew react be needing "use" react, server side views react i miss the sweet react, boot up from vite react i gotta say, at that time, i'd like to tweet react see i invented react, it wasn't any react and now i look and look around and there's so many reacts i used to love react, i used to love react i even had redux in it, i thought i was react what if react made an app about about react called it "i miss the old react"? man, that'd be so react that's all it was react, we still love react and i love you like theo loves react
102
102
1,804
410,354
Big true
26 Nov 2023
The growing complexity of software is not the worst thing. In fact it is completely natural, just as the growing entropy of a system. The worst thing is the people believing it is absolutely necessary and not only refusing to fight it, but actually defending it.
3
306
Apparently publishing usable types is hard There is a matrix of factors between all the various module resolution methods and ESM vs CJS This is a fairly specific comparison but hope this will save you some trouble in the future when, you too, have users
How to reference TS type definitions from package.json: github.com/codyebberson/type…
2
292
Proof that diffs in perf don't only come from the less performant code but the culture around the importance of perf or the lack thereof. I think Jarred's main contribution to the JS space is not Bun but the questioning of the status quo and asking if things really had to be slow
Replying to @ThePrimeagen
There’s a golden nugget in the ticket x.com/howardl3/status/170670…
2
235
I just realized I'm in a bubble "v8 engine" is a thing that executes JavaScript, but not to other people who spend time outside
26
49
707
118,330
Yeah I guess I'll check out this indie Next Conf thing or whatever nextjs.org/conf/ticket/thato…

2
346
Zustand is great, seriously underappreciated
13 Sep 2023
Let's celebrate 35k stars!
3
276
Seriously that always trips me up
I wonder how the guy who misspelled Referrer is doing.
2
251
Actually not a meme, good video 👍
12 Sep 2023
CSS margins scare me terribly, so I made a video about it youtu.be/KVQMoEFUee8
1
314
For ye who are brave Or like to go fast, either one
If you're a company migrating to Bun, fill this out We'll setup up a private channel for you on the Bun discord so we can help you more directly docs.google.com/forms/d/e/1F…
1
204
Matteo 👍 he gets it
This level of obsession with performance is why Bun is faster. @jarredsumner optimizes for the little things.. that could be hard to measure up. But they add up quickly.
1
218
Reminds me when I ran the Bun repo test suite for the first time It's literally game changing
11 Sep 2023
just yolo'd `bun test` on our monorepo jest test suite. I didn't expect everything to pass, so the result is pretty good. The speed is absurd 🤯
1
1
274

I like that this is a question (they’ve been using bun:test)
1
70
Pretty sure I've been annoying friends in the JS space by harping on this topic for years, but I think people start listening a little more when someone like Sunil, who publicly iterates quickly in spite of all the current obstacles in the ecosystem, says it. 100% this
Why is bun/biome/turbopack/<devtool with a focus on perf> _such_ a big deal? Why does a few seconds matter so much? As a product engineer, iteration speed is EVERYTHING. It literally multiplies runway. Do you see it? ime, In every non trivial project for the past 10 years, here's what's taken time to land code in CI (besides writing it): - install deps - build code - lint codebase - typecheck - run tests Mentally imagine one version of this takes 2 minutes, and other takes 20. Someone drops a review on your PR ("NIT", they say. Hmph.) So you commit it and wait again. Do you see the problem? Sure, you can throw caches at it, but beyond a point even loading/saving those take time (and money!!). We'd assumed our tools are as fast as they can get, so when product engineers, whose everyday work (read: do the above 10s of times from 9-5) come across a tool that reclaims their time, they associate it with FREEDOM. Do you see it? They can ship faster, spend more time with their family, get professional accolades, and focus on the things that help them ship. I'm not joking about this impact even a little. It helps them spend more time with their family. This isn't tech navel gazing. I'm speaking from experience. So bun comes along (for 1,2,5). Turbopack tries to be more holistic and makes a fast tool but also great caching. The bloomberg gang are going to solve 4. Biome might solve 3. And the impact is theoretically for _millions of developers_ across the world, since they're all working on conventions/standards. A PR that would take hours to land now takes minutes. Do you see it? It's a big deal.
1
192
Let's talk about how to identify a first edition Bun sticker Left is latest given out at 1.0 party. Right is 1st edition ultra rare Bun sticker, near mint condition. Notice the lighter color on the non-first edition. DM with offers, I don't get out of bed for less than 2k tho
3
392
Bun 1.0 you ask? Yeah I was there. I know, and no I'm not signing autographs at the moment. Yes I will open a PR to migrate your project to Bun though
1
17
1,173
Bun!
1
11
305
Yay! Congrats to the Oven team and all contributors on making Bun 1.0 happen. This is a milestone in JS history no doubt. A story we will tell our kids for generations... "Back in my day, we had to wait 40 minutes for our docs site to build, and 45 minutes for the tests to run!"
7
364