Hacker News Reader

Comment by trzmiel4 | Hacker News Reader

I've always loved Firefox. Had to switch to Chrome for stability and resource consumption, but I did not enjoy being all Google and I did miss numerous awesome features of Firefox. The awesomebar and dev tools to just name a few.

Went back to Firefox Quantum for a few months. Now I'm back on Chrome for exactly the same reasons. After hours of active use, Firefox is eating a ton of RAM and consuming way too much CPU, while also struggling to open any pages (let alone more complex apps) or downright crashing the apps that were already running.

Plus, it does not seem to support web calls at all (Google Meet, BlueJeans, GoToMeeting you name it).

I hate being on Chrome again, and I miss those Firefox features. But I have work to do, and I just can't have my tools grind to a halt all the time.

View on HN

I'm using all day Firefox for front develop, an I have multiple tabs open. I don't have any issue of excessive RAM or CPU usage. Even I noticed that Chrome uses more RAM that FF . The actual instance that I have open to write it, with 16 tabs open, it's using around 300-400MiB and ~5% of CPU.

Also, FireFox for Android runs faster that Chrome (not because FireFox is faster that Chrome on Android, but because the ad-blocker does a nice work avoiding to load garbage), and the RAM usage it's keep at bay. I have like ten tabs on FF for Android and keeps working like if I had one open.

On any case, on Chrome and Firefox you can install a add-on called "Tab suspender". It allows to save ram, literally suspending tabs that has not been accesses on some time. Try it.

*edit : grammar

> Even I noticed that Chrome uses more RAM that FF

Reflects my experience. I use Firefox at work, Chrome at home. Chrome RAM usage is much higher than FF. Not that I really mind - I have more than enough RAM to support it. But it often leaves me wondering if some tab/plugin/extension is misbehaving, which in turn forces me to close all my tabs only to find that the memory use stays high and the number of Chrome processes doesn't decrease as expected. I'm considering switching back to Firefox at home now.

EDIT: OK, tested with Firefox now. Number of process doesn't decrease as I close tabs either, but memory use is still considerably higher than Chrome.

In Firefox, entering about:performance in the URL bar will give you some statistics on how much memory or CPU each tab is using.

Entering about:about in the URL bar gives you a short list of interesting about:* links that you can play around.

I just saw that I'm using more RAM that I thought, but Tab suspender really works fine to keep ram usage at bay.

Firefox uses a process model that has a higher base memory usage than Chrome's but overall lower memory usage when used with a lot of tabs.

> I don't have any issue of excessive RAM or CPU usage.

I guess it depends per device and use-case. I use Firefox everywhere. Home (PC with Linux OS), work (laptop with Linux OS) and on my phone.

On my pc I have 0 RAM or CPU issues with Firefox. (intel i5, and nvidia GTS 250, ancient I know.)

On my laptop (intel i7 and nvidia NVS 5200M) I have had serious CPU issues though. With version 57 it used to stay above 90% CPU for hours. Since version 58 though it's been a lot better. Sometimes stays at about 40% for a while, then goes back down to like 10%. I suspect it might be due to some of the tabs I have open with zabbix graphs, but I don't know for certain.

I love using firefox on my phone because it's the only mobile browser that can use uBlock Origin. I don't care how fast or smooth chrome is on mobile, I'm not gonna use it if the price you have to pay is being eye-raped by ads.

the Brave browser is another alternative with ad blocking.

and its got smooth scrolling. (at least on android)

I'll try it out. Is the adblocker/s for it open-source?

i think so, but didn't audit it myself

* https://github.com/brave/browser-android-tabs

* https://github.com/brave/adblock-lists

* https://github.com/brave/ad-block

I'm also using Firefox on Android because of the Adblocker, but it is not faster than Chrome. The page load time is generally slower than Chrome, even with ads blocked. It sometimes freezes and crashes

I'm using a BQ Aquaris X with Android 7.1.1 . Zero issues with Firefox for Android.

I feel firefox on android in many ways far more better than chrome (especially the UX and addons), the only thing preventing me to make it my default browser is the built in page translation of chrome, for me its a must have since i live in a foreign country.

It makes me sad that there are a bunch of translation addons for Firefox but I have yet to find one that actually works on the Android version.

> I'm using all day Firefox for front develop

I want to do that but can't replace chrome's debugging features, specifically websocket network call analysis and hover variable to get its value. How have you replaced these feature in your workflow?

Here's what I do in such situations:

1. Record a profile with the Gecko Profiler WebExtension (Install it from https://perf-html.io/)

2. Upload the profile to https://perf-html.io/

3. Create a bug on https://bugzilla.mozilla.org/ with a link to this performance profile.

The Firefox developers are friendly and welcome, and I'd say in about 50-75% of cases the bugs get fixed in a timely manner.

Unfortunately being slow & hogging resources tends to more of a rule than an exception. I doubt it would help if everyone started drowning developers in bug reports that basically state the obvious.

It might also be that being the second fastest guy on the track will not win you a gold medal. As everyone tends to develop and test software on Blink/V8, it will undoubtedly feel slow when ran on a competiting browser.

I am a Firefox developer. Bug reports with specific examples of Firefox being slow, e.g. on a certain website, or with a certain hardware configuration, are very useful. Profiles even more so.

If it being slow really is a rule rather than an exception, then perhaps there is something in your configuration making it that way. Or a specific website that you happen to always have open. It's not normally always slow for most people, so again, this would be useful to know about.

It's true that often developers only test on Chrome, and in making sure it runs smoothly there they happen to make something that performs badly on Firefox. If we know about these websites then we can see if it's something we should change on our end, or get in touch with the developers ourselves.

I have been asking for this features for eternity, but now Firefox has changed a little may be there is a chance this might get consider again.

How about making this, Record, Profile, and Submission procedure built in. Button Accessible within 3-4 clicks. The amount of work required to get all these done and submitted is not easy especially for average users. I do a lot in my most enthusiastic days ( Mostly for Memshrink ), but after Memshrink I couldn't be bother much, unless the bug is annoying enough that persist.

> I doubt it would help if everyone started drowning developers in bug reports that basically state the obvious.

If the bugreport just reads "it's slow", then of course that's not helpful. However, performance issues as drastic as described above are certainly not normal. I'm using Firefox on Windows, Linux and (up until recently) macOS and never encountered anything remotely like this. So it would certainly be helpful to get detailed traces of that abnormal situation.

+1 for Firefox eating CPU. For some reason they can't track down the issue. If there's anyone from Mozilla reading, I'm on a Late 2013 13" rMBP fully loaded. I've made all the recommended changes to improve performance (and undone all of them as well to ensure they weren't the problem.) Firefox consumes most of my CPU and causes my fans to run at the full 6200 RPM while they try to cool my CPU which consistently reaches beyond 80C on any page running any Javascript.

Talking about "the issue" is a misunderstanding of how performance of complex software works. Overall performance is the result of the entire codebase, and there can be a varieety of ways that performance problems can manifest for individual users. One of the most productive approaches to improving performance with Quantum was fixing lots and lots of small and medium sized issues across the product that individually didn't have much impact but overall added up to a significant improvement.

There can be all sorts of reasons that specific sites or specific installations might show performance problems that aren't representative of the typical experience. Your specific environment might have some feature (particular hardware, profile data, addons, etc.) that happen to trigger a bad case in some code and so lead to an overall slowdown. Or you might be regularly using a site that happens to do something that's CPU intensive in Firefox but less so in other browsers. This could be a Firefox problem (maybe we implemented a feature using an algorithm with different tradeoffs compared to other browsers and it so happens that everyone avoids the worst case behaviour in Chrome, but not in Firefox, maybe our implementation just has issues that can be fixed), or it could be a site problem (sometimes sites just send buggy, broken code to specific browsers that causes them to use lots of CPU for no reason).

One of the things that was most valuable during the Quantum project was the work to improve the profiling tools. If you can capture a profile using the super-easy-to-use gecko profiler[1] and create a bug with the profile attached it should be possible to figure out what exactly is causing the problem in your case. Without that data it's really hard to make progress because the kind of problem you experience is just not something that would be allowed to ship if it was known.

[1] https://developer.mozilla.org/en-US/docs/Mozilla/Performance...

This. If it's becoming unusable on my pretty beefy i7/16GB (Ubuntu), how does it work for the average Joe?

It is a showstopper. I don't care about new features if it can't handle basic stability.

See, sibling post https://news.ycombinator.com/item?id=16647116 Please record a profile and report this.

Bad logic it works great on my several year old tablet and my 4 year old laptop with 2 cores.

If it was broken for everyone it would already have been fixed.

Works fine on my weak ass thinkpad (i3-4010U) o_0

Whereas on my i5 thinkpad ('14 t430) I had to switch to chrome as my browser, which was annoying, but even on a clean win10 insstall w/ only FF installed, it would peg the cpu hard. Don't know why, as I still use it as my primary browser for my desktop as well as android phone, but ever since Quantum there just seems to be a significant number of systems that are incompatible with firefox.

Oh, and before anyone asks, on the thinkpad the performance issue was still present on a debian stretch install as well.

Just because it works on your system doesn't mean this isn't a bug they need to fix.

One of the actual Firefox processes doing that? On my MBP the problem seems to be plugin-container, or most probably one of the third-party plugins running inside. Which is to say, either Cisco's H.264 codec or Google's video DRM module. `killall plugin-container` helps but also crashes basically all pages requiring video support, requiring reload.

Try opening about:performance from the address bar to see if a particular site is slowing it down

Even with no extension ?

I don't recognize your experience of Firefox in my own.

I use tree style tabs, so I have lots of tabs open - after a cleanup this morning, I still have 26, but 50 to 100 is far from unusual - and it's just as fast as ever to open new tabs. I don't notice the memory usage at all. Sure, browsers are all the biggest hogs - Chrome, Firefox, Slack - but with 16G in my work laptop there's no noticeable difference than with my home machine with 32G.

I don't use Chrome much beyond development testing, but with some features - anything with smooth animations in particular - I see way more CPU usage from it than FF.

I don't use Chrome much beyond development testing, but with some features - anything with smooth animations in particular - I see way more CPU usage from it than FF.

Just to add another data point, this has also been my experience: Firefox since Quantum often pegs CPU cores at 100% when showing pages with even quite simple animations.

The other big trouble area since Quantum seems to be extensions. After various experiments of my own and various discussions online, it seems clear that the machines I use where there are more extensions involved are much less stable and much more vulnerable to problems generally than those with mostly vanilla Firefox installations.

Yeah, one needs to check those extensions one by one for performance problems. My performance problems with Firefox went away when I uninstalled Ghostery a while back.

Firefox is doing a good job about analysing extensions startup performance, but I guess they need to do the same thing for page loads and memory usage. Or integrate the extension functionality into core more.

It seems to me that extensions used to be more integrated and the goal of the new model is, as much as anything, to separate and encapsulate them more. In theory, that should help with security and stability issues.

Unfortunately in practice what's happened so far is that lots of useful little things aren't possible any more because the new API doesn't support them, while extensions are evidently still capable of causing problems with performance and stability anyway.

Hopefully both of these aspects will improve in time. In particular, the emphasis on personalisation and improving the extensions ecosystem in this roadmap is welcome and seems very much in keeping with what a lot of us used to like about Firefox. I hope they really do put their focus there and exercise caution on becoming too "opinionated".

FWIW Firefox does support WebRTC (which is used for web calling), however sites like Meet sometimes filter for Chrome anyway (IIRC Meet may use Chrome-only proprietary features?). Hangouts is now available on Firefox though, after years of waiting for Google to do it.

How do you have hangouts on firefox again? Is it enterprise hangouts (pre-meet) or consumer hangouts?

Might be enterprise.

That hasn't been my experience at all since Quantum.

Using Firefox at work, often with more than 100 Tabs and multiple windows open without any issues. And my machine only has 6GB RAM...

Weird, Firefox Quantum greatly improved all resource consumption for me. Maybe you’re running into a bug.

It seems like the problem is that Google Meet doesn’t support Firefox. Not the other way around. Google’s unwillingness to make Meet work in all browsers really annoys me.

As much as I love FF, this is not true. Chrome supports VP9, FF/Safari do not. VP9 is a huge leap ahead of VP8, and Google is not wrong to use it in Meet. What's more, FF only supports UDP which requires a much less firewalled environment.

On another note, the lack of support for sessions is the real killer for me. And the lack of WebSocket debugging is a big strike against the dev in me. I use FF as my primary browser, but gotta give credit where it is due.

FF definitely supports VP9 (at least on Linux). Easy to check by loading up a YouTube video and right clicking to get the "Stats for nerds".

It supports VP9, but not in WebRTC. Or at least, not without enabling a flag on each viewers computer[0]. I am running Nightly, and can verify that it is _still_ behind a flag.

I suspect they don't realize how many viewers this costs them, as this has been behind a flag for months already, meanwhile RTC users are being told "We recommend Chrome".

[0] https://stackoverflow.com/a/38701342/87520

Google should support H.264 as well. The quality is close to VP9 (i.e. most users won’t notice a difference) and the hardware acceleration makes a huge difference on all but the latest systems. Saving a little bandwidth at the expense of going from 10% to 100+% CPU is a net loss, especially for battery powered devices.

Systems from last few years have VP9 encoding too, only AMD is a laggard in this regard (Raven Ridge is the first GPU to support it).

And while we are talking about hardware, it would be nice, if both Chrome and Firefox supported hw accelerated decoding on Linux. Currently, they don't support it at all.

Yes, it’s appearing but it’s not anywhere near a given, especially on corporate and non-gamer systems — exactly the majority of users who don’t care about bandwidth if it means their CPU fans are on high driving a slideshow.

I've been using Firefox as my main browser for years (typically with 30-50 tabs open at any given time) and I never once saw it use more RAM or CPU than Chrome.

Are you running a lot of extensions?

I wanted to love Quantum (really liked the Containers for cookies and sessions, found it useful many times)

but it takes up too much CPU. I kept track of the issue https://bugzilla.mozilla.org/show_bug.cgi?id=1404042

It seems it's just normal that every page render takes 80-90% cpu. I'll take battery life over marginal performance improvements.

Hope they'll make it a priority one day.

> Plus, it does not seem to support web calls at all (Google Meet, BlueJeans, GoToMeeting you name it)

I had many issues with Firefox and Webex. In the end just put it down to Linux or Ghostery plugin rather than Firefox specifically, as I couldn't get it to work with Chromium either.

However, in my experience Firefox does work flawlessly with appear.in

If it's a plugin or such, it may be helpful to make a profile just for that site (about:profiles or -P on command line). I do this for various sites that need different extensions or just that I want to isolate for privacy/security.

Unfortunately, performance doesn't come free on a computer. Have you tried limiting the number of CPUs Firefox uses?

You can see a setting "Performance" in the first tab, uncheck the "Use recommended.." and reduce the number of CPUs it suggest. Maybe that'll help?

I had the opposite experience, some years ago I switched from chrome to ff because certain recursive calls in javascript would lock up the processor and put my computer into an state where manual reboot was the only option.

I had similar issues, but found it was my TreeStyleTabs extensions consuming all of my CPU. When I disabled the extension the issue went away. Could it be an extension that was causing issues?

>I've always loved Firefox. Had to switch to Chrome for stability and resource consumption

Have you tried brave browser ? It was created by Brendan Eich, the co-founder of Mozilla and I'm so happy with Brave browser right now, especially in terms of resource consumption, it's just so slick and fast

Just another mild Chromium fork. What's the point?

>mild Chromium fork

Mild ? For some reasons, it's like 5X faster then chrome on my laptop, Can't live without it anymore

Is all this gain due to the adblocker?

Probably due to the very efficient scripts + ads blocker

For me, Firefox only takes more RAM/CPU when waking up from sleep. On a fresh reboot of Firefox, it always takes less CPU and RAM than Chrome.

At least it works for you, for me Firefox just starts eating gigabytes of RAM when starting up and then finally it stops and maybe becomes usable, but even then, without proper HW acceleration Youtube, WebGL games and kerning are awful.

If anyone has suggestions how I could get all those things fixed it would be really welcome. I'm using an Intel Core 2 6700 with Intel 965Q graphics with 8GB of RAM.

I can’t see your graphics card in the blocklist[1] so I’m not sure why hardware acceleration would be disabled. If you look in about:support is it definitely not working? If you go to about:config and search for ’acceleration’ is anything disabled there?

If none of that works, you could try going back to about:support and click on Refresh Firefox. You’ll lose settings and extensions but it can clear up problematic issues. [2]

[1] https://wiki.mozilla.org/Blocklisting/Blocked_Graphics_Drive...

[2] https://support.mozilla.org/en-US/kb/refresh-firefox-reset-a...

See: https://news.ycombinator.com/item?id=16647116

Please record a profile and report this.

Had the same experience. Give Opera a shot, I'm very satisfied with it since switching from ff a month ago.

Have you tried tab suspender?

Agreed. No mention of the severe memory issues. I like it, but my computer needs to be usable too.

I loved Firefox until it lost the plot. They tried to look like a clone of chrome. They removed the status bar because it was "superfluous" and then they integrated the pocket extension. I hope they regain the joy that it caused when I first used it

[Edit] Why the downvotes? I want Firefox to be successful, it just seems to have lost its charm from when it was originally called Phoenix[/Edit]

> Why the downvotes? I want Firefox to be successful, it just seems to have lost its charm from when it was originally called Phoenix

Not sure about the downvotes, either, because I think you have a very good point.

However, I suspect that many people want Firefox to be exactly like Chrome, at least subconsciously. This is of course silly, because what do we win if Firefox becomes a second Chrome? Also, it would mean that Firefox will always lose, because every little difference to Chrome would be considered negative.

On the other hand, it might make tactically sense for Firefox to mimic Chrome for now, to win back the lost users.

In the end, I want Firefox to be clearly different from and better than Chrome.

It's also fair to say that some people subconsciously go well out of their way to treat every vague similarity as part of some broader conspiratorial case that Firefox is trying to clone some other browser in general. Even when the products still ultimately look, feel and behave so differently that it's clearly not the case.

In fact it's painfully obvious that Mozilla isn't at all trying to "clone" Chrome, given that their recent Photon UI design looks less like Chrome than their previous Australis UI, all things considered. Not to mention that even their own Chromium-based products are drastically differently from Chrome, and they've just confirmed in this post that they intend to even move those over to GeckoView.

As such it's a very silly and tiresome argument to hear, and since the rest of the comment didn't seem very interested in having a conversation or adding to the greater discussion, I'm not surprised it would just get downvoted.

I wanted to love Firefox Quantum. At first, it was great. But for whatever reason (likely CPU and RAM as people are describing in this thread), I've discovered that its performance doesn't scale with many tabs as well as Chrome. I've discovered also that when I switch from tab to tab, sometimes the tab needs to reload the page. That's fine. But sometimes, it takes forever to get past a blank white screen on that tab. Sometimes when I open a new tab and try a Google search from the URL bar, same thing. Tried the same URL or same Google search from the URL bar in a new tab in Chrome, absolutely no problem.

Sorry, I'm currently switching back to Chrome. :(

> I've discovered also that when I switch from tab to tab, sometimes the tab needs to reload the page.

I believe that should only happen for tabs restored during startup. Unlike Chrome (unless something has changed), Firefox lazy-loads restored tabs on startup. I think that's preferable to the alternative.

Sure, but I said that's fine. My point is that it doesn't load. I wait for minutes, and it doesn't load. Quoting myself:

I've discovered also that when I switch from tab to tab, sometimes the tab needs to reload the page. That's fine. But sometimes, it takes forever to get past a blank white screen on that tab.

Reply on news.ycombinator.com