No announcement yet.

Firefox 109 vs. Chrome 109 Browser Benchmarks On Ubuntu Linux + Core i9 13900K

  • Filter
  • Time
  • Show
Clear All
new posts

  • #41
    Originally posted by Michael View Post

    They are just using selenium for triggering the web benchmarks... Selenium itself isn't the benchmark or measured as part of the tests.
    any chance you see the news about igalia being hired/sponsored or whatever it's called to work on servo? igalia will have 4 devs working on it


    • #42
      Originally posted by stormcrow View Post

      Don't need to check the source code. You can monitor the network connections it makes and which options trigger which network connections. Even if you notice an encrypted connection deep packet inspection will likely tell you what's going on. If you really want to know the contents and using a MitM proxy doesn't tell you the contents, you CAN read the source code for the function associated with it. This is a red herring of an argument. You can't read Chrome's source code and there's no way to know if any given Chrome release ties to any given Chromium release, or what the binary modules vanilla Chromium builds import are actually doing without trying to disassemble them.

      In short, it's EASIER to audit Firefox than Chrome if I and others so desire.

      Apply a little logic next time you try to refute a solid argument.
      I need research not vapid accusations.
      Last edited by avis; 22 January 2023, 07:46 PM.


      • #43
        Originally posted by kozman View Post

        Microsoft stopped trying to swim upstream with an anchor tied around its waist and just adopted a Chrome base and let Google continue to do all the hard work and heavy lifting. And thus reaping the metrics and ad revenue while lifting nary a finger. From a biz standpoint, it's a smart move. We all know Microsoft can afford to but it's always better to work smarter, not harder. Do next to nothing, make money and increase your own browser usage for free? Who doesn't want that revenue stream? You don't buy the cow when you get the milk for free.

        I suppose Microsoft can or does pass along all the hardware info and related crash / perf metrics to the Chrome code base because not doing so makes Edge (and thus Chrome) a shittier product on whatever various gear it's running on. Microsoft wins. Google wins. LAMF. I think FF on Linux isn't up to par with Windows or Mac because I think that's where their bread and butter comes from right now. From reading Phoronix articles and being a Mozilla bug tracker volunteer I read a lot of the bugs around Linux and it's improving. Sometimes glacially, but it's improving from where it was even 5+ years ago.

        Some of those issues with FF and Linux that were long thought to be unfixable or could never be worked around have, to some degree, been fixed or worked around. Some smartypants figured it out eons later but some progress is better than what has felt like zero progress. There will always be more to do but the good work is built atop other's contribution so that's always a win for the Linux community. A lot of improvements and fixes to the kernel, GPU drivers and MESA has helped move the needle in areas that were lacking or hobbling furthering FF. Compared to a decade ago when a lot of us were just shaking our heads at what Mozilla management was doing to FF, they seem like they've pulled their head somewhat out of their asses. Time will tell.
        I agree that Firefox ten years ago worked a lot worse than it does now and many Linux specific bugs have been addressed. Still under Windows it runs considerably faster and it's about everything: launching, rendering pages, opening new tabs. I've used Linux exclusively for over two decades now and sometimes I just wanna break my laptop 'cause how badly Firefox runs on it.
        Last edited by avis; 22 January 2023, 07:34 PM.


        • #44
          Originally posted by shanedav4 View Post
          Video performance with Chrome on Linux is horrible. Open two Chrome windows and play a web video in each one. Now try moving the windows around while the videos are playing. You'll find that the performance is horrid.

          Firefox on the other hand has none of this problem and performs smoothly.
          Why would you move a browser window at all? The vast majority of people simply have their web browser maximized at all times.

          And I'm sorry to break it to you, but Chrome is more power efficient at rendering videos than Firefox when both use software decoding. I'm talking about the official applications. Chrome wins by a large margin considering my power meter and fans. Don't know why but WebRender is super ... shaky under Linux. Maybe it was designed for Windows in the first place.


          • #45
            Originally posted by Ahmad_S792 View Post
            Michael - I would personally like to see WebKit Epiphany in the mix as well.

            I'll save you both the trouble. It bogs down HEAVILY with multiple tabs in use, to the point that it's not usable on current hardware with more than about 6 tabs open. Try it.


            • #46
              Originally posted by caligula View Post

              OTOH modern laptops support tons of memory. 2 x 32 GB DDR5 + 90% ZRAM @ zstd helps a lot if you need dozens of active tabs.
              I'm not so sure you actually know the laptop market nowadays. Sure, laptops with that much RAM do exist, but they usually cost insane amounts of money. The vast majority of devices can be lucky if they ship with more than 8 GB. That's true for both Europe and northern America. Sure, ZRAM helps, especially in combination with zstd, but 90 % ZRAM is just insane and I can't imagine how it should not be slowing down your work majorly as the system is constantly compressing and decompressing stuff. And I'm not sure how efficient that can even be done.

              Adding to that, this is simply no excuse to write terrible software that wastes memory for nothing. It only makes software even worse on any not maxed-out hardware.

              Also, this doesn't say anything about power usage. Even more, such an amount of ZRAM is guaranteed to draw more power as more progressing time has to be wasted to compress and decompress RAM.

              To sum it all up, your comment is just a bad excuse for bad software written by lazy developers expecting ideal circumstances (huge amounts of memory, bleeding fast CPU and RAM, plugged to the wall) in any use case. But throwing more resources at a problem instead of simply fixing it and investing in "human resources" is always a bad idea.


              • #47
                Originally posted by avis View Post

                Have you read and checked the entire Firefox source code? Has anyone here on Phoronix done that? No? Then how can you claim Firefox is not spyware?

                Secondly, have you reverse engineered Google Chrome? Then what on Earth allows you to baselessly claim that Chrome is/contains spyware?

                God, most vocal Linux/Open Source users are so embarrassing it's just cringe.
                It's been widely studied and reported on in an overwhelming amount of academic research. For example, read the research by Douglas Leith from Trinity College - "Web Browser Privacy: What Do Browsers Say When They Phone Home?" Chrome phones home each letter you type, as you type it, uniquely identifies your system, and can track you across browser restarts and browser re-installations.


                • #48
                  Originally posted by andyprough View Post

                  It's been widely studied and reported on in an overwhelming amount of academic research. For example, read the research by Douglas Leith from Trinity College - "Web Browser Privacy: What Do Browsers Say When They Phone Home?" Chrome phones home each letter you type, as you type it, uniquely identifies your system, and can track you across browser restarts and browser re-installations.
                  Sorry, I've just read it:

                  Chrome, Firefox and Safari all tag requests with identifiers that are linked to the browser instance (i.e. which persist across browser restarts but are reset upon a fresh browser install). All three share details of web pages visited with backend servers. This happens via the search autocomplete feature, which sends web addresses to backend servers in realtime as they are typed.

                  1 Chrome tags these web addresses with a persistent identifier that allows them to be linked together. Safari uses an ephemeral identifier while Firefox sends no identifiers alongside the web addresses. The search autocomplete functionality can be disabled by users, but in all three browsers is silently enabled by default. Chrome sets a persistent cookie on first startup that is transmitted to Google upon browser restart

                  2 Firefox includes identifiers in its telemetry transmissions to Mozilla that are used to link these over time. Telemetry can be disabled, but again is silently enabled by default. Firefox also maintains an open websocket for push notifications that is linked to a unique identifier and so potentially can also be used for tracking and which cannot be easily disabled.

                  3 Safari defaults to a choice of start page that prefetches pages from multiple third parties (Facebook, Twitter etc, sites not well known for being privacy friendly) and so potentially allows them to load pages containing identifiers into the browser cache. Start page aside, Safari otherwise made no extraneous network connections and transmitted no persistent identifiers, but allied iCloud processes did make connections containing identifiers.

                  In summary, Chrome, Firefox and Safari can all be configured to be more private but this requires user knowledge (since intrusive settings are silently enabled) and active intervention to adjust setting

                  Yeah, Firefox is so much better.


                  • #49
                    Originally posted by Artim View Post

                    The first part might be mainly because of more sane defaults in the flags. But for the bold part you really have to explain that. Last time I checked, Chromium Edge was so bad at privacy protection, Chrome looked like a saint. They send everything back to Microsoft, every single keystroke, even in instances where even Google doesn't bother to or decides at random when to send what information.
                    I didnt even really consider telemetry and was mostly looking at this in the past:

                    But that looks fairly awful now too, lmao.

                    I never even imagined Edge as a privacy friendly browser though, hence it always (at the very least) has to be supplemented with something else.


                    • #50
                      Originally posted by avis View Post
                      Firefox also maintains an open websocket for push notifications that is linked to a unique identifier and so potentially can also be used for tracking and which cannot be easily disabled.

                      Yeah, Firefox is so much better.
                      Well, as a volunteer for Mozilla for 13+ years, that push notification thing is often used to push out a GPU blocklist (for users still rocking ancient hardware) or setting revert/disablement (for that new feature they wanted to enable in release but it turns out it breaks something so has to be rolled back or disabled quick) in the event that some new issue crops up and requires an immediate fix rather than having to push out a new dot-release version. It's not some evil cloak and dagger operation TBH. And you can always disable it in about:config if you know which setting it is or where to look. There's no free lunch. FF did suck on Android but 102 onward is leaps and bounds better but for sure not as fast as Chrome. When I had my Nexus 5X using Chrome, every damn second some random bullshit page hijack event was always happening on the most benign sites. After installing FF 70 series and onward with uBlock Origin, never had one single hijack event anymore. For all it's slowness and faults on Android, the BS stopped and I've stuck with FF and all the growing pains on Android. Not the fastest but I have zero issues when used with uBlock Origin.

                      Remember when you used to be able to disable "OK Google" with a setting? Because it was Alexa always-listening-creepy? And then, mysteriously, that disablement setting disappeared with whatever Android version? Everyone knew it was listening and spying and disabled it but Google just said fuck it, leave it on and remove the ability to disable it. Google Assistant, even disabled, still listens. I think all companies do some manner of this 'cause the data analytics bros ain't cheap yo.