Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
×
Chrome Firefox Chromium Opera Linux

Which Linux Browser Is The Fastest? (zdnet.com) 160

ZDNet's Networking blog calls Firefox "the default web browser for most Linux distributions" and "easily the most popular Linux web browser" (with 51.7% of the vote in a recent survey by LinuxQuestions, followed by Chrome with 15.67%). But is it the fastest? An anonymous reader writes: ZDNet's Networking blog just ran speed tests on seven modern browsers -- Firefox, Chrome, Chromium, Opera (which is also built on Chromium), GNOME Web (formerly Epiphany), and Vivaldi (an open-source fork of the old Opera code for power-users). They subjected each browser to the JavaScript test suites JetStream, Kraken, and Octane, as well as reaction speed-testing by Speedometer and scenarios from WebXPRT, adding one final test for compliance with the HTML5 standard.

The results? Firefox emerged "far above" the other browsers for the everyday tasks measured by WebXPRT, but ranked near the bottom in all of the other tests. "Taken all-in-all, I think Linux users should look to Chrome for their web browser use," concludes ZDNet's contributing editor. "When it's not the fastest, it's close to being the speediest. Firefox, more often than not, really isn't that fast. Of the rest, Opera does reasonably well. Then, Chromium and Vivaldi are still worth looking at. Gnome Web, however, especially with its dreadful HTML 5 compatibility, doesn't merit much attention."

The article also reports some formerly popular Linux browsers are no longer being maintained, linking to a KDE forum discussion that concludes that Konqueror and Rekonq "are both more or less dead."
This discussion has been archived. No new comments can be posted.

Which Linux Browser Is The Fastest?

Comments Filter:
  • X11 vs the world (Score:3, Interesting)

    by leenks ( 906881 ) on Saturday March 04, 2017 @07:51PM (#53977875)
    One of these browsers, given it's history, is most likely optimised for X over other screen rendering platforms. Sure, the benchmarks concentrate on JavaScript, but ultimately this is going to result in manipulation of the DOM in any real world test and therefore re-rendering... Does this therefore boil down to how well these browsers optimise their performance on X, as opposed to anything else? Given the changing landscape (Wayland and alternatives) does this have much relevance?
    • by dbIII ( 701233 )
      Just use something without gtk3 and you'll have something optimised for X.
    • by Anonymous Coward

      I've yet to find a benchmark that shows Wayland as a winner on speed. So far X11 is faster on almost all tests I've read.

    • by Z00L00K ( 682162 )

      And what is really the issue considering the fact that most of the slowness on the web is either the network or the remote server.

      • by AmiMoJo ( 196126 )

        We need to get beyond simple "race" style benchmarks. The guys reviewing GPUs for gaming realized that years ago and now look at things like the 99th percentile framerate and micro-stuttering.

        For a browser "speed" as experienced by the user is a combination of things. Rendering performance, network performance, tricks like pre-loading and speculative DNS lookups, JS performance... But also the UI and how quickly it allows things to be done, how responsive it is. A great example is Internet Explorer/Edge. Do

    • by tibit ( 1762298 )

      X is largely irrelevant. Today, X is used to do three things:

      1. Push pixmaps from the application to the screen. Notice that nowhere does X get involved in doing any rendering of those pixmaps.

      2. Push UI events from the user to the application, and poorly participate in window management.

      3. Allow applications to open a windowed OpenGl context.

      It's a lot of cruft that does nothing much, only so that some obsolete pure X11 application will still keep on working. Architecturally speaking, it's nonsense. For mo

  • by Mr307 ( 49185 ) on Saturday March 04, 2017 @07:55PM (#53977895)

    Once robust standards were being followed and browser speed went past some point, I stopped caring about which was the fastest. Care much more about interface features and plugins that I want. Next was the many other annoying things that I was able to customize to my taste, a menu item up or down on a list, a button I could or could not move, maintaining a familiar interface, etc.

    • Unfortunately for me FF is still a bit laggy, but it is still the best browser overall. Vivaldi is the next best and is definitely snappier that FF especially with a resource-intensive site like Twitter. Vivaldi seems a lot faster for sites with video. I generally use both. I've used Konqueror but yes it looks like there's basically not much going on with it and it doesn't even have a lot of basic options one expects in a browser. Opera is ok but I had it blow up on me not long ago. I think Vivaldi is

    • by mysticgoat ( 582871 ) on Sunday March 05, 2017 @03:27AM (#53978969) Homepage Journal

      I've been using FF since it was new. I have occasionally looked at other browsers and several are faster than my FF, at least partially because the plugins and modifications that I use slow FF down somewhat. But FF is fast enough that changing to a faster browser would not improve my productivity. And I've got a nice set of plugins and extensions on it that I would have to put together from scratch if I changed browsers. That is, assuming other browsers offered similar features, which as near as I can tell, they do not.

      Speed isn't the only criterion for measuring a browser's goodness. The ability to tailor it to your personal workload is much more important these days. And once you've got a browser tweaked to your best practices, do you really need to take the massive hit of finding, installing, and configuring the plugins of some other browser that would duplicate what you've already set up in your old FF?

      If you really need a faster browser, most of us who have been around the block would be better off running the same browser and OS on faster hardware. But this doesn't apply to young'uns who have yet to establish productive work habits. Their best approach would be to talk with some older guy who knows what he is doing about which browser he uses, how he has it set up, and what his workflow is.

      • It's amazing how fast a browser actually can run once you ban certain script elements from loading, certain calls and manipulations of the DOM (I don't think it's the business of most websites to check if I have a microphone or webcam installed and enabled nor what type and brand of hard disk I am using, nor do I think they have the right to disable any of my input devices or functions thereof for any reason), and all advertising. Even the "slow" browsers run amazingly well compared to what we had to choose

    • by tgv ( 254536 )

      Precisely. Who cares about the last drops of speed? If a page is really heavily Javascripted, like some of those data browsers, then JS speed matters, but otherwise only plugins like Block, NoScript and Disconnect plugins are a very good reason to stay with Firefox. Chrome is also more of a memory monster.

  • by msauve ( 701917 ) on Saturday March 04, 2017 @07:55PM (#53977899)
    Lynx. It doesn't have to deal with all those bandwidth intensive graphics.
    • Unless your output goes to a text-to-speech reader, elinks is worlds better. Lynx doesn't even have tabs.

      But indeed, both win over graphical bloatware.

      • Hasn't elinks been abandoned? The last unstable update was 5 years ago, and the last stable version was 3 years before that. Out of curiosity, I just checked the website and almost all the links to user forums, dev forums, etc. are dead. Lynx and Links (the latter having some features of elinks) are at least actively maintained.
    • by lucm ( 889690 )

      Lynx. It doesn't have to deal with all those bandwidth intensive graphics.

      That's why I always use data URI in my web pages instead of images. You don't have to look at my graphics, but you're gonna download them anyways.

      • A black-and-white image? convert trollface.png trollface.ubrl && cat trollface.ubrl
        A colour image? catimg [github.com] -r2 | ansi2html (package colorized-logs), elinks with use_document_colors=2 only.
        A histogram? braillegraph [github.com].

        The first and the last work in plain Unicode text, the second one requires HTML.
        On any site with basic Unicode support I'd include samples, but, you know, Slashdot...

  • by Anonymous Coward

    Chrome is Google product. Doesn't it continually report to the mother-ship? That would make *you* the product.

    • People say that but have yet to produce any packet logs showing what it does. So until someone steps forward with proof I will chalk this up to neckbeard paranoia.

      • It keeps coming up because we'd like to understand why Google would spend money developing a web browser.

        • I think it would be obvious why Google would want to provide a free web browser to its products (some might mistakenly call them "users"). They want to ensure their products have safe, secure, and functional access to their services, and controlling the browser means they get to control that experience to a larger degree.

          But I highly doubt Chrome spies on its products. Why? Because there's no need for the browser itself to snoop when 3rd party cookies do the job just as well "legitimately". A huge perce

          • Well, if you happen to sign in to Chrome (which I do) to use their sync service (because I and my household are all bought into the Android ecosystem), Google flat out admits that unless you encrypt your data with a passphrase so that they can't read it, they will parse any data you keep stored on their servers to sync to your other devices.

            The downside to using a passphrase, is that if you accidentally forget what you used (which in my older age, has happened once now), then your only option is to clear th

        • Google derives income by connecting the publishers of documents on the web with sponsors through services such as AdSense/AdWords and DoubleClick. When people find documents through Google Search, use other Google properties, or view a participating publisher's documents, Google gets a cut of the ad revenue. The more web browsing, the more money for Google. So a better browser is likely to keep people on the web longer.

          Internet Explorer stayed a piece of shit for far too long. Then Microsoft actually got a

        • Why does Microsoft spend money on Edge?

      • by tgv ( 254536 ) on Sunday March 05, 2017 @03:42AM (#53979001) Journal

        I run LittleSnitch on my mac, and when I open Chrome, it calls to several of Google's addresses: appspot-preview, gstatic, www, and fonts.. It also does that when opening an empty tab. It gets images and fonts and whathaveyou from those sites, (all unnecessary, BTW: the page functions just as well when all traffic is blocked), and it of course reports URLs for "malware" detection. That should give Google a nice bunch of data to work on.

    • I just block all outgoing traffic to the internet. Then chrome cant communicate with motherbrain
  • by zedaroca ( 3630525 ) on Saturday March 04, 2017 @08:02PM (#53977919)

    Said a marketer who does not understand what Linux users expect from their computing.

  • by marcel_in_ca ( 1850606 ) on Saturday March 04, 2017 @08:09PM (#53977947)
    Without AdBlock or NoScript, most websites are unusable to me.
    • by Baleet ( 4705757 )
      Reading this on Chrome, AdBlock tells me it has blocked 53,805 ads since installed.
  • by gosand ( 234100 ) on Saturday March 04, 2017 @08:14PM (#53977957)

    I switched to it a few months ago from FF, and it seems much more responsive to me. It is especially better in startup-to-response time, where FF was taking 30 seconds.(no, I didn't have a ton of add-ons or customizations)

    Speed really is only one piece of the puzzle. I was satisfied with the speed of chromium when I tried it for a while, but FF has the features I use. I much prefer the way FF does bookmarks, the bookmark toolbar, and tabs. That is why I have been very satisfied with Pale Moon... the features of FF I need without the bloat and dog-slow response.

    • by Wolfrider ( 856 )

      --I've been using Palemoon for the last couple of years now (mainly on Linux, but also on the Win side occasionally) and it definitely uses less memory and crashes less often than Firefox in my experience.

      --Personally I don't care which browser is the fastest, PM has been meeting my needs and is more STABLE.

      • That wasn't my experience at all. It would crash a lot more, and there were pages that would render fine in Firefox which would render incorrectly in PM. So I gave up and went back to Firefox, which now has 64 bit builds.

    • by Doke ( 23992 )
      Pale Moon 27.1.2 got "381 out of 555 points" at https://html5test.com/ [html5test.com].
    • If any application is taking 30 seconds to load up, you've got issues. Heck, LibreOffice loads up a complex spreadsheet for me in less than 5 seconds.

      Maybe you're using an old hard drive that needs serious defragmenting? Or your configuration files for the app are totally borked?

      I get messed up configuration files slowing down my Banshee startup. I just blow away the config directory and it's good again for a few months.

      • by gosand ( 234100 )

        My hard drive is less than a year old. My install of Mint 18 was a fresh install. I tried a new FF profile, new config files. Plenty of RAM (8 GB, less than 1GB used)

        The issue was that FF would start up in the same time it always would, within a few seconds. However, it would sit and wouldn't take any kind of input for 30 seconds. CPU was quiet, RAM was quiet, disk was quiet. It even did this if I passed it a URL from the command line. It would sit for about 30 seconds, then would become responsive.

    • by AmiMoJo ( 196126 )

      Do you have a load of hidden tabs in Firefox? I was using the old Spaces thing or whatever it was called, and had about 50 tabs saved in there. Turns out that was making it take 30 seconds to start up. Deleted them all and it was down to 1 second to become responsive again.

      It was the UI buggery that made me move to PaleMoon.

  • by POWRSURG ( 755318 ) on Saturday March 04, 2017 @08:55PM (#53978085) Homepage
    Vivaldi is not an open source fork of the old Opera code base. That code base was proprietary. It is blink based. The confusion is that it was founded by the former Opera co-founder/CEO, and aimed to restore features Opera lost when they moved to Blink. It's more a fork in the Opera consumer base and is not a fork of the code base.
    • by AmiMoJo ( 196126 )

      Vivaldi is actually a fork of Chromium. Their code base tracks Chromium, with their own custom changes to add in Opera like features. The UI is identical to Chrome in many places.

  • Dillo or W3M are the fastest web browsers. Dillo is a GUI web browser with only HTML/PHP (not sure about ASPX) & CSS, SSL support, cookies (change Dillo config file), and a simple wget GUI interface as its download manager. It does incredibly well for going to websites you don't need to view videos on. I even got it running on my Mac (couldn't get SSL support to compile). W3M is a web browser that runs in a command line and some terminals (not all) support w3m-img so even images can be shown as the back
    • Never mind about Chromium. My bad.
    • Pale Moon browser? It's basically the same as Firefox before the 64-bit eye-candy nazis got ahold of it. It's really fast too. 'Member?ðY (South Park joke).
    • by fnj ( 64210 )

      PHP and ASP are server side. Yeah, they can use GET and POST, but they are not necessary to the use GET and POST. Any HTTP client "supports" HTTP generated by PHP and ASP.

      BTW, SSL support in dillo is only an alpha prototype plugin. It does no cert caching or authentication.

      • They will have SSL built in by 3.1 though, rather than as a plugin; though, not sure how advanced it will be. I think the dpi info for it is somewhere in here: http://www.dillo.org/dpi1.html [dillo.org]. In all fairness, I can't get --enable-ssl to work on my Macbook, but that may be an OpenSSL issue. So, I've got HTTPS and HTTP working on Linux, but no HTTPS on my Mac. It's got a few bugs, but it opens instantly and works perfect for looking up information really quick. Using Ctrl+s to search Duckduckgo has issues; yo
  • by koavf ( 1099649 )
    There are six browsers, not seven.
  • "WebXPRT: This is today's most comprehensive browser benchmark. It uses scenarios created to mirror everyday tasks."
    Basically, Firefox is not the fastest at all things Javascript synthetic performance, but it's the fastest for real world web-browsing.

    Based on that... I would actually recommend using Firefox?

  • I can guarantee that if you compile it yourself, with optimized CFLAGS and CXXFLAGS, that Firefox and Chromium will come out ahead of the poorly optimized builds that get released by the non-open source browsers.

    Personally I use compiled versions on my Funtoo [funtoo.org] Laptop and Workstation. Yes they take some time to compile but if you only upgrade twice a year it's not so bad. Upgrading every release would simply be too much. My CFLAGS are nothing crazy: "-march=native -Os -pipe"

    There is also the added benefit tha

  • by Bronster ( 13157 ) <slashdot@brong.net> on Saturday March 04, 2017 @10:06PM (#53978337) Homepage

    The summary is wrong that Vivaldi is a fork of old Opera (the Presto engine), it is in fact the same Blink engine that powers Chrome and new Opera, but with brand new chrome (non-capital, aka the interface around the engine) which is recreating the power-user features of old Opera rather than the cut-down interfaces that other browsers are working towards:

    https://en.wikipedia.org/wiki/... [wikipedia.org]

    https://en.wikipedia.org/wiki/... [wikipedia.org]

    https://en.wikipedia.org/wiki/... [wikipedia.org]

  • Konqueror (Score:4, Insightful)

    by StormReaver ( 59959 ) on Saturday March 04, 2017 @11:24PM (#53978555)

    As a Web browser, Konqueror has sucked for a long time. While there was a time when I used Konqueror exclusively, its renderer fell increasingly further behind the other browsers. It needs to die as a Web browser.

    As a file manager, though, it is still top dog. Its killer feature (aside from I/O Slaves, which are awesome) is the ability to split the screen into multiple panels. Dolphin is a brain disease that needs to die, just for its inability to do more than two panels, and Konqueror as just a file manager needs to resume its rightful place as the default file manager.

    A second-best would be to have Dolphin be able to split panes like Konqueror does. At least then, Dolphin would no longer be the abomination that it is.

    • Why is split panes better than simply being able to do Windows 7-like window snapping? It's not like you can't afford the memory for another window, right? Then the feature doesn't have to be in the app at all.

  • by markdavis ( 642305 ) on Sunday March 05, 2017 @12:02AM (#53978635)

    >""Taken all-in-all, I think Linux users should look to Chrome for their web browser use," concludes ZDNet's contributing editor. "

    And maybe ZDNet doesn't understand Linux users. Many, perhaps even most of us, do not want a closed-source, closed-developed, semi-spyware, anti-configuration-friendly Chrome browser as our preferred browser.

    Oh, and his main benchmarks: Speedometer is a "webkit-designed benchmark" and it surprises him that the webkit based browsers did considerably better than Firefox (the only non-webkit browser in his lineup)? Then a Google (think webkit again) based javascript benchmark, same result. Yet when he used Kraken and Jetstream, miraculously the browsers were just a 10% and 11% spread (with Chrome not winning either). BTW- he never ran Ooort, which Firefox seems to always win (and by a lot), and Peacekeeper, which Firefox usually wins.

    My take: Firefox does just fine with speed. It is not the fastest, but the speed difference isn't as much as one might think, and it certainly isn't the only important factor when choosing a browser.

    • Many, perhaps even most

      This is confirmation bias. Linux is getting more and more popular by the day, and a large portion of the user base couldn't give a shit about proprietary vs open debate as long as it fits their use cases.

      A lot of people switched to Linux because it's not MS and they didn't want to buy Apple hardware.

  • Comment removed based on user account deletion
  • by newbie_fantod ( 514871 ) on Sunday March 05, 2017 @02:28AM (#53978891)

    But who cares. It's security/privacy that I want. All the browsers seem just fine speed-wise.

  • You'll need a fast CPU anyway (Core 2 Duo and up), gigs of RAM and the patience to live without H264 hardware acceleration if your graphics are too old or unsupported (such as Radeon 5450 under Ubuntu 16.04).

    There's no other way around if feeding your computer with mounds of garbage to be parsed, interpreted, compiled and encabulated.
    Fun fact : didn't PDF files use to be slower than web pages? Now it's the other way around.
    Maybe we could make a "website" out of a PDF document with hyperlinks to other PDF fi

  • by Anonymous Coward

    Trump Supporters have small genitalia. Why?

  • Memory usage, not speed is the killer in modern web browsers. People often have many windows open, and each page could suck up a gig of memory, the system rapidly becomes unstable. Let's see some tests about memory and stability with large numbers of open pages.
  • I use Chrome across all devices, because it enables me to take advantage of the Google-Android ecosystem, which is its main advantage for me.
  • I notice browser slowness much more on my ARM devices than even on Celerons. I think the biggest differentiator is who can multi-thread the best to take advantage of available processors.

  • Actually it doesn't matter which one is the fastest.
    Browsers are optimizing since years and with each release they claim to be 10-30% faster. If it were true, pages would be rendered instantly ... ... and on the other hand, nobody selects browsers by rendering speed. People select them by interface, startup time and addons. This matters a lot more than if the page renders 10ms faster.

  • On JavaScript benchmarks, all the browsers are within 5%. Nothing that the end user will really notice.

    The only test where Firefox falls flat was designed by the web-kit team, and suprise, suprise all the other browsers listed are web-kit variants.

    Really though servo is set to knock all the other browsers right out of thier socks.

  • Firefox is really fast, depending on many factors ...

    I upgraded from Kubuntu 14.04 to 16.04, and Firefox is the fastest I have seen in years. Maybe it is because Canonical have switched to the 64bit version, or something else. This is on an 8 year old laptop.

    Regardless, you have to install the following:

    - uBlock Origin (block ads)
    - NoScript (Disable Javascript and Flash except from white listed sites)
    - Cookie Monster (Don't accept cookies except from white listed sites)
    - Classic Theme Restorer (Makes it loo

news: gotcha

Working...