Results 918 comments of dan

Here's what Claude had to say about the root cause based on a crash dump from my phone and looking at React and Webkit source. Crash dump: [com.apple.WebKit.WebContent-2025-12-09-203451.log](https://github.com/user-attachments/files/24055100/com.apple.WebKit.WebContent-2025-12-09-203451.log) Stack trace:...

Speculative fix from Claude https://github.com/facebook/react/pull/35337

Non-React repro if someone wants to file a radar: ```html ::view-transition-old(*) { animation: 300ms both fade-out; } ::view-transition-new(*) { animation: 300ms both fade-in; } @keyframes fade-out { to { opacity:...

Yea not sure, I guess let's try it and see if I can hit it again at some other codepath

not gonna do right now but please remind me to look next week

While I know how to write a debounce, is there any benefit to this? The Lodash one works fine.

I think it's fine not to do this.

We need to consider Japanese and a few other locales that don't use 3/6/9/12 as the breaking points.

Here's what ChatGPT told me: >Where Factor of 10^3 Doesn’t Work >Here’s a summary of locales where the standard K (10^3) and M (10^6) approach doesn’t apply directly: >- East...

right, all of the PRs we've received so far on this topic don't deal with these locales — i'd take a fix that does