Browser Engines

December 3rd, 2021
While there are many web browsers, they're almost all wrappers around the only three actively developed full engines. Those are:

You could also make the case for two others:

  • Flow: A clean-slate closed-source browser and engine, designed to make effective use of parallelism. This matters most on highly resource-constrained devices, and if you have a Raspberry Pi you can download and play with their demo. The reason I am not including this engine above (yet!) is that while it's made impressive progress, including handling Google Docs, it's still a ways from done; the demo page has "we're not quite ready to start receiving reports for rendering issues just yet, these are mostly down to missing features, so as we add them you'll see those issues steadily disappear over time." You can see which features they have and haven't implemented on the web-platform-tests dashboard. To get a very rough sense of where it stands today we can compare the proportion of web platform tests that pass in each browser:

    Chrome 97%
    Firefox 96%
    Safari 94%
    Flow 24%
  • Goanna / UXP: A pre-Quantum fork of Gecko, developed for Pale Moon. There doesn't seem to be an official SCM: links to a snapshot archive and an archived GitHub repo. It looks like it was on until sometime between May and October 2021 (wayback). Probably ~2021-08-25, since that's the final public commit. Based on the timing, probably deleted it as part of the MyPal licensing blowup? They don't seem to participate in web-platform-tests. This looks like a fork that has fallen well behind Firefox (ex: no web components or nullish coalescing), and I expect we'll see them fall further behind and either shut down or rebase on top of one of the actively maintained engines.

I've heard some concern that we're losing browser engine diversity as the web platform grows and it becomes harder to build and maintain something that supports it fully. While I do think the size of the web platform could have that effect, I'm not seeing it yet. Some of this is that there just haven't ever been that many competitive browser engines. In the late 90s, we had three (powering Netscape, IE, and Opera) and then four (KHTML) in the early 2000s. In 2003 Safari (WebKit) forked from KHTML, making five. For three months in 2013 we had six, between Chrome forking WebKit to make Blink (April) and Opera discontinuing Presto for Blink (July). In 2016 KHTML development fell off, with consumers switching to first WebKit and then Blink, bringing us down to four. In 2020 Edge switched from their own engine to Blink, bringing us down to three. And now with Flow we'll probably be back to four soon.

Comment via: facebook, lesswrong

Recent posts on blogs I like:

A Big Problem With The Going To Bed Book

One day my dad was reading this book called the "Going to Bed Book" to my sister Nora. The book is basically about a bunch of animals who are getting ready for bed on a boat. They go down the stairs, take a bath, hang their towels on the wall, find…

via Lily Wise's Blog Posts September 18, 2023

Investing in boundaries with young kids

Putting in some work to get the behavior you want The post Investing in boundaries with young kids appeared first on Otherwise.

via Otherwise August 15, 2023

Self-driving car bets

This month I lost a bunch of bets. Back in early 2016 I bet at even odds that self-driving ride sharing would be available in 10 US cities by July 2023. Then I made similar bets a dozen times because everyone disagreed with me. The first deployment to pot…

via The sideways view July 29, 2023

more     (via openring)