A new CEO for Mozilla

It does if you want users. If you want users, you need to support 99.99% of websites.

And even with a simpler browser engine, you're still going to get bug reports and support requests. The only way you're going to support a browser engine with three developers is by not handling much of the web *and* not having many users.

> The things they add to the browser engine aren't being used by users. They're being used by developers.

Who develop things used by users; that's always how it works when providing a platform. Some of those features *are* quite user-visible. And while some of them are primarily for developers, they help replace developer hacks with better solutions, or put the user more in control, or otherwise do things that developers are seeking a way to do.

It used to be much hackier to produce certain types of page layouts; now it's easy, in any modern browser. There are a lot of features that used to require JavaScript that can now be done with pure CSS or HTML. Web developers don't tend to say "I can't do that", they say "it's going to be harder to do that", and then they make it happen, whatever that takes; browsers can make hard things easier and simpler. Browsers can reduce the size that webpages need to be, or improve page compression and load time, or improve security...

By way of an example: I'm not a fan of webpage-specified fonts, but the world *before* they existed often had sites render key text as *images*, making it harder to scale, worse quality, larger, less accessible, and unselectable. Now, in a modern site, you don't see people put text in an image just so they can use a particular font, and that's a good thing. It'd be an even *better* world if websites didn't do *either* and just used browser fonts, but 1) nobody has succeeded or is likely to succeed in telling web developers what they *shouldn't* do, and 2) you can configure your browser to ignore webpage-specified fonts, which is one way browsers keep you in control, and they can't do that with the text in an image.

> Users just want websites and basic web apps.

You may want to examine substantially more broadly than the bubble of users you are in. Users want games (both casual and fancy), high-quality videoconferencing with screen sharing, sites that load quickly and securely, sites that work on mobile devices with limited/unreliable bandwidth while providing nice features, sites that include real-time communication with other users, and many many many other things. Pretty much any functionality you could want a computer to do, users want to be able to do on the web.

> Much of the new things that get added to these browsers are actually bad for users for god's sake, like tracking features.

Tracking features are absolutely a bad thing that have no place on the web. The vast majority of new features are not inherently bad in the way that tracking is.

In fact, some of the new functionality being added to web browsers is to *improve* privacy, such as locking down third-party cookies and similar tracking mechanisms, and otherwise locking down bits of the web platform to put users more in control.

One of the oldest features on the web is showing visited links in a different color than unvisited links. There's a huge amount of infrastructure that has gone into preserving that functionality in browsers while not allowing that to let pages track whether you've visited a page or not. Firefox has a whole mechanism to limit what sites can change between visited/unvisited links, so that there's no way for a site to *infer* that you've visited or not visited any given link.

Here's an example. I have a few Bafang mid-drive motors on my cargo bikes. There are a bunch of useful settings that you can change, and much like a customizable linux desktop, I usually don't like the stock settings that the vendor programmed. It used to be that you needed some (free as in beer) windows program to update the settings, but someone reverse-engineered the protocol and people have written various (free as in freedom) programs to update settings (eg, https://github.com/horga83/bdac). That's all fine and good, but I really didn't want to have to build/install random software that I'll only use once every 5-10 years when I get a new motor. Turns out someone wrote *a website* that can update the motor settings. Which is amazing. As simple as just plugging my laptop into the bike, going to a website, turning on the motor, and punching in the new settings that I want. I don't have to mess around with containers and host permission to ttyS*, or pulling down 50 python dependencies and realizing that something was written in python2 but never updated for python3 and debian no longer supports python3, and and and.. I just go to https://devnotes.kymatica.com/BafangWebConfig/BafangWebCf... and do the thing.

But that website works in chromium but not firefox, because it "requires a browser with support for Web Serial API, and should work on any recent version of Chrome, Opera or Edge." https://developer.mozilla.org/en-US/docs/Web/API/Web_Seri... says that it was added with chrome 89, which would've been mid-2021.