This benchmark draws a very different picture having GPT5.5 on the very top with 70% and DeepSeek at 8%

https://deepswe.datacurve.ai

DeepSWE has been heavily criticized though. https://github.com/datacurve-ai/deep-swe/issues/21 Putting GPT 5.5 on top is the obviously correct part, but everything else about it makes very little sense.

I get motion sickness from this specific effect. Especially on high refresh rate screens.

Thats why I maintain the successor to tachyons: https://tachyonsneo.com No build pipeline. Resort to CSS when utilites make no sense. No lock in.

Can you clarify how this relates to Tachyons - do all Tachyons features work? Or is it a subset of Tachyons?

It's just a rebuilt drop-in replacement stylesheet compatible to the built version of tachhyons 4.12 but using css variables (so when you need to write custom css you can use its tokens) and it has support for grid plus a few more modern utilities.

Building an application you use this as your utility layer that gets you 80% and for the rest you use an application specific custom css to write the parts where css is the best.

I've been yak shaving my little rework of tachyons for quite some time now. I like to drop it into projects and embrace the narrower design constraints it gives me. If it can be of help for anyone or you have ideas how a modern tiny css only utility framework should look like let me know.

Hopefully Mythos didn't go rogue and hold production hostage.

Today is a good day.

They should have just scaled a proper Rails monolith instead of this React, Java whatever mixed mess. But hey probably Microslop is vibecoding everything to Rust now!

Team is doing resume driven development

They are „Rust“ing here: https://starship.rs/

I’ve seen Bezos’s yacht and can tell you he is creating new jobs! His yacht is accompanied by a supply vessel that carries staff, food, etc. At least 20 people work on the secondary boat!