Benchmarking Firefox 83 Nightly With "Warp" Against Google Chrome On Linux

[Michael_S]

The page to look at is the bottom half of https://arewefastyet.com, which benchmarks the page load times of popular websites using common settings (that is, a Facebook home page for someone with hundreds of connections, an Amazon page for someone with an extensive purchase history, a Twitter page for someone that follows and is followed by many other accounts, a Google Doc with a lot of content, and so forth). Benchmark information here: https://wiki.mozilla.org/TestEnginee...or-tp6-1_to_10

On those benchmarks, Firefox stable is as fast or faster than Chrome on: Amazon, Apple, Ebay, Facebook, Fandom, Google (!), GMail (!), Google Slides (!), Google Docs (!), Imgur, Instagram, LinkedIn, Microsoft, Paypal, Pinterest, Reddit, Tumblr, Twitter, Wikipedia, Yahoo Mail, Youtube (!), and Yandex.

Chrome beats Firefox on IMDB, Bing, and most of the synthetic benchmarks.

All this tells you is that most of the synthetic benchmarks are useless for measuring real world user experience.

Mozilla has a lot of serious problems, starting with a 2.4 million dollar salary for the CEO of a non-profit that just laid off 250 staff. But Firefox performance isn't one of the problems.

[xnor]

All this tells you is that most of the synthetic benchmarks are useless for measuring real world user experience.

Mozilla has a lot of serious problems, starting with a 2.4 million dollar salary for the CEO of a non-profit that just laid off 250 staff. But Firefox performance isn't one of the problems.

But that is simply not true. First of all, what is really measured here? Is it the time to load a page 100%? Then this tells you nothing about user experience.
Total page load time could be 200% and yet the browser could feel way more responsive, could start rendering much earlier, etc.

[xnor]

Secondly, I cannot post or paste the rest of my reply because the forum is broken.

[blackshard]

Come on, this is not serious. I mean, benchmarking alpha-state software, probably with debugging flags/code active, and then comparing with stable too...
Come on, this is not professional at all

[ferry]

but i do

Yeah, I feel bad for you.

But, my point is, you will always find problems. How bad is your web site compatibility compared to (Google) chromium locking up the OS when using Google Earth? The former can be solved by firing up Chromium, the latter by a hard reset

[Mez']

Secondly, I cannot post or paste the rest of my reply because the forum is broken.

Are you using Chrome?

[Mez']

> Get funded by Soros Open Society
> Ultra-liberal hotheads get appointed as managers and chiefs
> Tranny developers start pouring in
> Trannies push good developers away from project
> Project starts going to shit
> Even Soros' money is not enough
> Must fire trannies even
> Only remaining dudes are just pajeets in internship
> Tries to regain market by "going social" with ultra-liberal propaganda (usually asking to censor freedom of speech and pushing mainstream fakenews)
> Must fire even last bit of developers
> Make a release with SciFi nomenclature, only to find it slowed down shit even more

Is this how browsers die?

I don't like the way you envision Firefox's future.
But I really like the non political correctness. Especially how supposedly tolerant and democratic ultra-liberal tend to end up dictating their own single version of the truth and censoring freedom of speech when meeting opinions not fitting their narrow views.

[AmericanLocomotive]

The page to look at is the bottom half of https://arewefastyet.com, which benchmarks the page load times of popular websites using common settings (that is, a Facebook home page for someone with hundreds of connections, an Amazon page for someone with an extensive purchase history, a Twitter page for someone that follows and is followed by many other accounts, a Google Doc with a lot of content, and so forth). Benchmark information here: https://wiki.mozilla.org/TestEnginee...or-tp6-1_to_10

On those benchmarks, Firefox stable is as fast or faster than Chrome on: Amazon, Apple, Ebay, Facebook, Fandom, Google (!), GMail (!), Google Slides (!), Google Docs (!), Imgur, Instagram, LinkedIn, Microsoft, Paypal, Pinterest, Reddit, Tumblr, Twitter, Wikipedia, Yahoo Mail, Youtube (!), and Yandex.

Chrome beats Firefox on IMDB, Bing, and most of the synthetic benchmarks.

All this tells you is that most of the synthetic benchmarks are useless for measuring real world user experience.

Mozilla has a lot of serious problems, starting with a 2.4 million dollar salary for the CEO of a non-profit that just laid off 250 staff. But Firefox performance isn't one of the problems.

I was going to say, I think a lot of these synthetic browser benchmarks don't really tell the story. A user is not going to notice a difference between a 250ms and a 300ms page load.

I have a pretty modest internet connection, and everything loads near instantly on FireFox. Granted I have a near TOTL CPU (Ryzen 3900X), but these days even many mobile CPUs match or exceed it in single threaded performance. Imgur, G-Mail, Google Slides, Sheets, Docs, Youtube load fast and quickly. I never sit there going "oh man, I wish this would render faster". It's basically near instant.

Firefox going to Chromium/Blink would be a disaster for the web, IMO. You'd basically be allowing a single company to completely shape and dictate how the web should look and work. It'd be as terrible when IE was the "defacto" browser. It'd take just one person to find a zero-day Chromium exploit and cripple everything.

[elatllat]

Both Firefox and Chromium are buggy, glad I have 2 FOSS options... links -g is sometimes a light option but so much of the web is needlessly incompatible and bloated.

[Michael_S]

This has been beat to death, but to be clear: Mozilla Foundation had, until the layoffs, 1,000 employees and their "Outreachy" diversity initiative covered 20 interns for six months per year. Since it's a public non-profit their demographics are public: 75% male, 75% white, and the percentage of men and whites in management is even higher. I don't think they publish "trans" demographics, but their percentage of employees in 2019 that identified as neither men nor women was 0.2%.

Mozilla leadership made lots of serious mistakes over the past twelve years, but blaming the browser's problems on their diversity programs is like claiming the Titanic sank because it had too many gay bartenders.

But that is simply not true. First of all, what is really measured here? Is it the time to load a page 100%? Then this tells you nothing about user experience.
Total page load time could be 200% and yet the browser could feel way more responsive, could start rendering much earlier, etc.

But the other benchmarks - Ares, Speedometer, MotionMark - don't measure those things either. It seems to me that "time to full page load", imperfect as that may be, is more useful than "runs a set of Javascript and DOM manipulations in a combination totally unrelated to sites human beings actually visit in a browser."