ScriptedAlchemy's profile picture. Infra Architect @ByteDanceTalk. Maintainer of @webpack @rspack_dev - creator of #ModuleFederation #auADHD #synesthesia own opinions.

Chief Wiggum

@ScriptedAlchemy

Infra Architect @ByteDanceTalk. Maintainer of @webpack @rspack_dev - creator of #ModuleFederation #auADHD #synesthesia own opinions.

Thanks ChatGPT!!

ScriptedAlchemy's tweet image. Thanks ChatGPT!!

Chief Wiggum hat repostet

If you just heard about lynx-ui from @ReactAdvanced , try our demo here yourself! QRCode in the comments!

Introducing L.U.N.A Lynx UI New Aesthetics



Chief Wiggum hat repostet

In the next release, Rspack will generate a more compact module factory format that trims the wrapper function. Benchmarks show this reduces bundle size by about 1% before gzip, though the effect after gzip is minimal. 😉

rspack_dev's tweet image. In the next release, Rspack will generate a more compact module factory format that trims the wrapper function. 

Benchmarks show this reduces bundle size by about 1% before gzip, though the effect after gzip is minimal. 😉

Chief Wiggum hat repostet

Imagine dragging your app onto a drop-zone and… nothing else. No picking clouds, no YAML yoga, no “oops, forgot the VPC.” We call that Superpositioned Infrastructure. Think of it like this: your code is a bunch of couch-surfing friends who never bother renting their own place.…


I have rsc working with federation. Two variations. 1) http proxy (kind like vercels rsc federation) 2) full capabilities of rsc like nesting and server actions which involve tracking all the refs across apps etc. 3) wip: single server using “code pull” similar to node mf


Chief Wiggum hat repostet

In the next version of Bun, Rsbuild and Rspack work thanks to N-API compatibility improvements

nektro's tweet image. In the next version of Bun,

Rsbuild and Rspack work thanks to N-API compatibility improvements

Chief Wiggum hat repostet
jait_chen's tweet image.

In the next version of Bun, Rsbuild and Rspack work thanks to N-API compatibility improvements

nektro's tweet image. In the next version of Bun,

Rsbuild and Rspack work thanks to N-API compatibility improvements


Chief Wiggum hat repostet

Rspress v2 RC is live across all Rstack sites. 🎉 A few fixes are in progress. The final release will ship once everything is polished.

rspack_dev's tweet image. Rspress v2 RC is live across all Rstack sites. 🎉

A few fixes are in progress. The final release will ship once everything is polished.

I haven’t encountered a single problem that throwing a billion tokens at didn’t resolve. Cost issues are temporary, and can be subsidized anyways.


Can’t wait for “GPT-5.1-trippin-balls-400-mics”


Chief Wiggum hat repostet

Just tried @Trae_ai. Gotta say, it's the most underrated IDE. People are sleeping on this.


Code is a side effect of the prompt. As long as the AI can maintain it and understand it. I could care less, chances are I’ll never edit another file by hand again.

the hardest lesson i learned in my career was that code quality doesn't matter. i realized this while being nosy and checking out dynamodb's source code. simultaneously one of the most reliable, rock-solid AWS services and the worst codebase i have ever seen



Chief Wiggum hat repostet

Built with @LynxJS_org

We’re launching a new Time & Wellbeing page and mindful tools designed to help our community pause, reset, and build positive digital habits on TikTok. #DigitalWellbeing



🚨 GPT-5.1-Codex-Max just dropped! - #1 on SWE-Bench → 77.9% - 24h+ long-running tasks (actually insane) And it can handle much bigger codebases thanks to compaction, which condenses its history and keeps only the key context We have a new SOTA SWE model

flavioAd's tweet image. 🚨 GPT-5.1-Codex-Max just dropped!

- #1 on SWE-Bench → 77.9%
- 24h+ long-running tasks (actually insane)

And it can handle much bigger codebases thanks to compaction, which condenses its history and keeps only the key context

We have a new SOTA SWE model
flavioAd's tweet image. 🚨 GPT-5.1-Codex-Max just dropped!

- #1 on SWE-Bench → 77.9%
- 24h+ long-running tasks (actually insane)

And it can handle much bigger codebases thanks to compaction, which condenses its history and keeps only the key context

We have a new SOTA SWE model
flavioAd's tweet image. 🚨 GPT-5.1-Codex-Max just dropped!

- #1 on SWE-Bench → 77.9%
- 24h+ long-running tasks (actually insane)

And it can handle much bigger codebases thanks to compaction, which condenses its history and keeps only the key context

We have a new SOTA SWE model


Loading...

Something went wrong.


Something went wrong.