Floati Browser / Release notes

v0.1.0 - 0.1.9 · 10-12 Jul 2026334 releasesv0.1.343 · 6 Oct 2026

147 built 79 underway 79 still on the roadmap 0 lit by this release

0The engine exists

1Reading-grade rendering

2App-grade JavaScript & web APIs

3Media

4Graphics, compositing & the GPU

5Security & multi-process

6The rest of the web platform

7Accessibility, tooling & polish

★The reference browser parity, on our own engine

Hover or drag along the releases above to see the engine at that moment. Each feature lights up in the release that started it and the release that finished it; recent releases also light every feature they worked on.

Browse releases by topic

v0.1.32428 Sep 2026·Rendering, JavaScript & JIT

Long nested loops run to the end, and reading long text is fast

Two fixes found while building the next JavaScript slice. A nested loop that did a lot of work could stop with an error saying the script ran too long after only a few milliseconds, and reading long text one character at a time slowed down the longer the text was. Both are fixed on every tier.

Parity with the reference browser

74overall, as of v0.1.343

Web APIs / DOM surface42

Security architecture66

GPU / graphics acceleration68

DevTools / extensions72

Networking85

Media85

Accessibility85

Real-world site compatibility99

HTML parsing100

CSS parsing + selectors100

Layout: block / flex / grid / tables / writing-modes100

Paint + visual100

JavaScript language100

JavaScript performance (JIT)100

HTML parsing100

What changed 2 changes

JavaScript & JIT	Long nested loops run to the end. Compiled loops pause about every millisecond to check whether you navigated away; an inner loop that was mid-way through at that moment was treated as a script that had run too long, so ordinary code such as 50 rows of 100,000 items stopped with an error. The inner loop now checks in and carries on, and only a script that really runs for 30 seconds is stopped.
Rendering	Reading long text character by character is fast. charAt, charCodeAt, codePointAt and length used to copy and rescan the whole text on every call, so a script walking a long text took time proportional to the square of its length. Long strings now keep a small index of their characters: walking 60,000 characters of accented text went from 3.7 seconds to 16 milliseconds, and plain text from 383 to 13 milliseconds.