It could be a platform difference. I mostly use macOS; sounds like you're on Windows.

I use FF on macOS and Linux; a bit on gaming Windows machine as well. On macOS it's about the same as Chrome, on Linux Chrome is comically horrible, even 1080p YouTube videos stutter (using GPU acceleration). Developing using Chrome under Linux (Ubuntu) was making everything super slow, e.g. debugging some JavaScript or GWT code was reducing performance of web app comparing to FF like 5x, making it unbearable.

Yes, it has been quite widely reported here and elsewhere such as the Firefox bug trackers, that the Firefox performance on Mac is quite inferior. Which makes me very sad. Though it does not seem to affect all Mac users

Firefox integrates into some platforms better than others. There are some proposed changes to fix this, but they aren't quick fixes.

>Write bindings to all relevant OS composition libraries. For Windows, we already have DirectComposition bindings, and for Linux, there are already bindings to Wayland. We need bindings for macOS Core Animation (which runs in the window server on relatively recent versions of macOS) and Android SurfaceFlinger.

https://github.com/servo/webrender/issues/3115

On the Mac you can pick up performance improvements now by opting out of the OS's window scaling feature for the Firefox app.

Get info for the Firefox app in the Finder and select the option to "Open in Low Resolution".

Count me as one of the people for whom Firefox runs butter smooth on my Mac. Never slows down or gets janky.

Very strange how there is such wide discrepancies in experiences for Mac users.

If they want to gain market-share they should prioritize Macs, since most web developers these days prefer them. The more developers who use Firefox, the more it'll get tested with, the better support it'll have.

My understanding from reading the bug reports is that the general cause of macOS performance issues is understood, but to do anything about it with the current architecture would require pulling developers off WebRender development instead. Instead, work is being done to make things better there for when that's ready to ship on Mac.

Is there any way to provide incentives to devs to prioritize mac development? Like a bug bounty of sorts?

I'd happily contribute to a patreon or something if it meant increasing the chances of having a usable version of firefox on my mac at some point.

To think that Mozilla can have revenue of $500M+ per year, yet a Patreon might be needed to make their browser faster :-)

The problem in osx is not that it's slow, but the ridiculous amount of cpu (and therefore battery) it consumes.

For the people who suffer the bugs ( which seems to be a majority of users that have a retina display, and specifically those who use it set on high resolution) just opening firefox will quickly lead to the macbook's fans spinning at full throttle and battery drain. There are issues opened and development supposedly in progress, but it's been like that for months if not years. At this point I would take whatever could speed up the process...

If Firefox is chewing through all your CPU from the moment it opens you need to reset your profile.

https://support.mozilla.org/en-US/kb/profile-manager-create-...

I don't think Firefox can fix all the bugs at the same pace Chromium can.

I've uninstalled and installed Firefox several times during the years, in different machines even (MacBooks between 2013 and current models). Same feeling from most developers working at my workplace (~1000 devs in the building, not one using FF on a Mac ).

For some reason every time I talk about these bugs on the internet people treat it as a weird edge case I'm going through, but talking IRL with people about Firefox everyone seems to know how unusable it is in a Mac - not only in my circle of acquaintances.

Reinstalling Firefox on a Mac is not the same as doing a Firefox Refresh. The former only touches the executable while the latter cleans up cruft in your Firefox data (potentially lingering around from legacy extensions).

I understand that, but I assume both of those are still "shallower" ways of resetting than actually switching computers, which I've done several times (for unrelated reasons ofc) and I've seen no improvement...

Is that true outside of US?

I'm not sure. But the benefit it provides is a Linux-ish environment that you don't have to tinker with constantly to keep it running, which is a potent combination. And web devs don't care about fine-grained hardware control/compatibility.

You know, I wonder if it's GPU-bound. Macs often have weaker GPUs than Windows machines, and this could explain the inconsistent results people are mentioning in the thread.

You may be hitting this: https://bugzilla.mozilla.org/show_bug.cgi?id=1429522

Firefox does not use Core Animation, unlike Chrome and Safari, which causes an extra full-window blit on GPU in some cases. It also uses a transparent window, so macOS has to composite the content behind the window whenever Firefox draws. I have a patch in there to improve things, as well as a crate planeshift to help the situation with WebRender, but the patch causes some usability regressions so it has not landed. In the meantime you can set gfx.compositor.glcontext.opaque to true in about:config to get an opaque window at the cost of rounded window corners and vibrancy.

I highly doubt this is the case. I can't tell the difference between my gaming desktop and my laptop with integrated graphics.

From what I've seen on the chromium bug tracker, Google has been on a ~5 year crusade to improve Mac performance. Around half the issues Ive seen include something like "We don't know what Mac does here, let's ask some people at the next conference" (these were core rendering performance bugs). I totally get why Firefox can't compete with that. It's fast on the platforms that don't seem to be hell-bent on making custom rendering slow.

Ah-hah!

"What if we stopped trying to guess what layers we need? What if we removed this boundary between painting and compositing and just went back to painting every pixel on every frame?

This may sound like a ridiculous idea, but it actually has some precedent. Modern day video games repaint every pixel, and they maintain 60 frames per second more reliably than browsers do. And they do it in an unexpected way… instead of creating these invalidation rectangles and layers to minimize what they need to paint, they just repaint the whole screen.

Wouldn’t rendering a web page like that be way slower?

If we paint on the CPU, it would be. But GPUs are designed to make this work."

https://hacks.mozilla.org/2017/10/the-whole-web-at-maximum-f...

Are you being sarcastic? I can't tell.

Any modern CPU and GPU can do a lot of DMA and copying of memory in a very short amount of time, but that doesn't mean it'll be efficient wrt. power... which is what I think you were 'complaining' about?

The GPUs can do it, push 60 page updates per second, but that is very energy inefficient way to do it.

You can feel this by playing any 3D shooter on your laptop and see the battery being sucked empty.

Nope. I'm on a Mac with a powerful GPU and Firefox still sucks.

Some anecdata: I haven’t had performance issues with Firefox on Mac for a very long time. I’ve had more issues with Firefox on Windows than I have on Mac.

As for Chrome, I used to use it a lot, but the privacy concerns and controls it comes with bothered me too much to want to continue to use it on any of my systems.

Firefox containers are also essential to me while browsing the web - I don’t know if Chrome provides anything similar.

I find Firefox to be faster on both Windows and macOS. (No idea about Linux—I've heard it lacks hardware acceleration there, so that's probably bad.)

On GNU/Linux, Firefox is a much more responsible than Chromium. Chromium randomly decide to eat 100% of a CPU core time for no apparent reason.

I am on mac. I run all 3 browsers. Chrome feels the most sluggish, with firefox the clear winner.

Firefox seems to perform especially well compared to Chrome if you have lots of tabs open at once (dozens or hundreds). The only exception to that for me has been roll20.net, which is nearly unusable for me with Firefox but works fine in Chrome.

Windows user here and I agree with OP that Chrome is much faster than FF.

I have a really hard time understanding where this commenter is coming from saying FF > Chrome. Have not heard anyone else with that opinion.

I recently switched to Firefox on Linux after 6-7 years of Chrome and I do use Windows and Mac laptop too.

1. On Linux - Firefox feels much more snappier than Chrome. Specically on Fedora, Chrome as a input lag which is non-existant in Firefox.

2. On Windows and Mac - Chrome is either on-par or feels snappier. I use either browser on these on two platforms.

3. On Android (galaxy-s8+), Firefox is fine but it integrates poorly with rest of the OS. Youtube links don't open in youtube app, searching for locations is bit of pain because Chrome produces Google maps hyperlinks.

I am still sticking to firefox though, because my main computer is Linux and Firefox is just to much better on Linux (lately).

Point 3 is by design. An Android icon will appear in the address bar if a native app is available. A tap on it will delegate the current URL to the app.

On Linux, Firefox can now run natively on Wayland (you have to manually enable it) while Chromium still can't (but the support is being added). The responsiveness difference is massive

I'm going to be honest I've pretty much never understood the appeal of Chrome over Firefox aside from it running separate tabs in separate processes, but that stopped being a distinction forever ago. I find the Firefox UI better and performance to be equivalent in most cases, and in all the ones where it isn't Firefox was better.

And I have a really hard time understanding where the OP is coming from saying Chrome > FF ;)

It is odd; I don't dispute that Firefox evidently has some performance issues for some people – there are enough reports from people on varied enough hardware that I don't doubt the authenticity of what people are reporting.

But personally, for me Firefox is noticeably faster than Chrome, and I know I'm definitely not the only one.

For reference: I run Windows 10 (currently 1803) on i7-7700K/32GB with typically 20-30 tabs across multiple windows; Firefox RAM consumption can be up to 4GB after a whole day's use like that. I never find the browser slows down at all, and is always immediately snappy and responsive... whereas Chrome, in a way which is hard to quantify or describe, does feel as though it's a few milliseconds behind me sometimes.

I also find the same on my truly atrocious m3/4GB Surface though (massively regret that one, "should be alright for what I need" I thought as I reached the checkout, should have known better). Chrome quickly kills the Surface, Firefox manages to keep going far longer...

So in conclusion... I don't dispute that some people find Firefox to be slower, but I know for sure that isn't true of everyone. I have no complaints whatsoever about Firefox performance and have been using it daily for 18+ months (switched from Vivaldi... now there's a browser which has performance issues, for me at least...)

What kind of GPU does your windows machine have? I'm curious if that's the differentiator.

AMD R9 series something or other.

My hunch is the CPU is the differentiator. Chrome eats RAM, FF eats CPU. Maybe someone with more knowledge here has a more educated guess.

I did recently update my BIOS and it improved FF performance dramatically, reinforcing my belief it's the CPU. Pre and post spectre fix or whatever.

I'd like to know how heavy your browsing is in terms of the number of tabs and windows that are open. In my experience chrome blazes it's competition when it doesn't have as many tabs open since it inhales the RAM unlike Firefox whose default behaviour is lazy loading.

This also frees up my RAM for other memory intensive operations leading me to use Firefox more often.

Typical day, at the end of the day (machine has been up for 8+ hours) I'll have 20-30 tabs (or more) open across 5+ Firefox windows.

Some of those tabs will be lightweight, like GitLab issues, forums or webpages I'm building, whereas others will be more complex sites and web apps... Office 365 apps, YouTube, social networks, Figma instances etc.

Firefox will consume sometimes 4GB or more by the end of the day.

I run an i7-7700K and 32GB of RAM though, so it's never going to get near a low-memory situation. I'll end the day with about 50% total system memory consumption... no app is individually too demanding, just I run a ton of things simultaneously and because there's so much RAM available, Windows never has to take action to kill anything or reclaim memory back again.

This is particularly evident in UWP apps (although I hardly ever use them)... if I open one at 9am and close it, I can then reopen at 9pm and it will be there instantaneously despite effectively being cold start from user perspective, because it's been suspended in RAM all day!

Right now I'm sitting at 50.2% RAM consumption, of which Firefox is using 2.1 GB over 7 processes.

Next biggest are Figma (400 MB on a big document), Microsoft Teams, Spotify... oh yeah, Electron RAM usage is still abominable. But I don't need to worry about it ;)

agreed... firefox is fab for me as well.. the only issue that i have with it is sped up videos don't pitch correct well. they sound a bit strange and I can't listen at speeds higher than 2x.

For me, Chrome runs much smoother than Firefox on macOS with the 2013 MBP.

where devtools is concerned - do you use Firefox Developer?

I don't no, just regular F12.

Sadly, I still find Chrome to be much easier to work with for dev than Firefox. Firefox has made great progress lately with the devtools, and the new bits like Grid Inspector are awesome, but it's sad the basics still aren't there.

It's the little things... Firefox still displays cryptic JavaScript errors with internal stack traces, whereas Chrome gives clear explanations and always links the stack to a line in my code. That's just one example, none of the issues I have are anything major really in isolation, but together, I find Firefox less efficient to dev with than Chrome.

The sad thing... I really, really wish this wasn't the case, because it unfortunately reinforces the point made in the article.

Despite all my best intentions, I do end up building for Chrome first. I loathe and resent that but it saves me time overall. Every new release I try Firefox tools again though and I'm confident they're going in the right direction... hopefully in another few months, I'll be able to report back and confirm I use Firefox as my primary dev browser too.

Just to complete this info, not the usual "Dev Edition" referring to beta/alpha, but Developer Edition for web developers, on http://getfirebug.com