The Baseline Interpreter: A Faster JavaScript Interpreter in Firefox 70

It's at least "100% more evil" than other browsers.

Google doesn't need to plaster the world with Google Analytics if it can get most people to use a browser that phones home.

From around the time of the Windows 8 transition I used Microsoft Edge as much as possible. Firefox was at a low ebb then.

I switched back to Firefox when Microsoft announced it would use the Chromium rendering engine for Edge. At that point Firefox had improved performance a lot and I've mostly been happy with it.

The minus of it however is that many developers are choosing to only support Chrome. For instance I worked at a company that had developed a data analysis tool with a React front end and it didn't work with either Firefox or Edge (or Safari) so I had to install Chrome for work. I don't think there was a deep technical reason why that was, but rather they did not want to go through the effort to test on other browsers. Our customers weren't clamoring for wider browser compatibility so that was OK for the business.

From time to time I find public web pages that have problems w/ Firefox, although more frequently I find pages that don't like it that I block ads at the "hosts" level. Some sites now use trackers as part of the authentication/anti-fraud process and that can be a problem.

Believe it or not I hardly ever log into Google. I have a gmail account that I barely use, but when i do I IMAP into it with em client. I am really done with Adsense, Adwords, Analytics, and all that. If I am working for somebody that is using Google services I will use it, but otherwise I can go a month or two w/o logging into Google.

> Gmail was still dog-slow the last time I tried, Youtube can send my fan into a frenzy and Docs is also regularly problematic.

It seems to work fine on my system running Ubuntu 18.04 with ublock origin installed. Which OS are you using?

Chrome regularly tries to trick me into logging into my Google account via Chrome. If you try to disable this functionality, updates will include new dark patterns to trick you into logging in anyway. Now if I log into any Google service, Chrome will magically log into my Google account, too. When I open Chrome, I'm greeted with a new tab with a login screen for my Google account.

I don't want to have mess with a bunch of settings just to turn off anti-user features and tracking. I am willing to mess with a bunch of settings if it enhances the functionality of the application I'm using.

Open 'about:config' in Firefox and look at all of the ways Firefox's behavior can be configured. I run my own Firefox Sync instance, because Firefox is just that customizable. Google regularly removes customization options from Chrome. I used to be able to Cast non-HTTPS resources from Chrome, then I had to enable a setting buried in its experimental features to do so. Now the feature is "enabled" in the settings, but after Chrome auto-updated a few times, the feature doesn't work at all.

I can no longer install Chrome extensions from GitHub, even though I could a few weeks ago. Google decides to take extensions off of their Chrome Web Store, and then make it difficult to use extensions they don't approve of.

Firefox uses less memory and CPU than Chrome does, I'm on a MacBook, so anything that unnecessarily drains my battery is a pain to use.

Basically two reasons:

A lot of code is run once, so the initial compilation cost is potentially more expensive than just the process of compilation, without even running the generated code, and if not that then compilation+execution is also longer than just interpreting.

There are also benefits for performance. If you haven’t run any of the code yet (through the interpreter), the baseline portion of the JIT has no knowledge of the code that is running, and so has to make a bunch of static guesses and (if you look at the old baseline JIT in JSC) include a lot of code for various common cases that may happen. Eg in the JSC case every arithmetic op include the entire code for both integer and floating point arithmetic inline. That bites you in a few ways: the guessing type info means that you have poorly chosen branch ordering (eg you put integer logic in the middle of the “fast” path for something that is floating point), and you simply generate more code which itself takes more time to produce, and also results in increased icache pressure which hurts runtime performance.

I haven’t read the entire article yet, but the JSC interpreter is written in a pseudo assembly so it can do an on stack replacement to switch into jotted code more or less anywhere during execution as the wonders of assembly means you can guarantee identical layout.

Uhhh what? JSC beats v8 across a wide variety of code, and has stable ABI so you don’t have to have N different copies of the entire implementation across every app.