Announcement

Collapse
No announcement yet.

Power & Memory Usage Of GNOME, KDE, LXDE & Xfce

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • THE FIVE STAGES OF BENCHMARK LOSS

    Stage 1: SHOCK

    The first reaction when you see an article, scanning down to see your competition wipe the floor with you.
    Typical postings: 'WTF', 'no way', 'you lie'...


    Originally posted by Adarion
    What the f.?
    Originally posted by energyman
    [...] the whole thing is complete bull.
    Originally posted by KAMIKAZOW
    What's that bulls* battery usage test? Do those numbers [...] even mean anything [...]. If you feel the need to benchmark Linux DEs, at least do it in an objective and transparent way...
    Originally posted by aavci
    DO NOT USE UBUNTU TO BENCHMARK MEMORY.
    Originally posted by SkyHiRider
    [...]who knows what garbage does Canonical bundle with both the desktops.
    Originally posted by kraftman
    Thanks for your FUD, Phoronix
    Stage 2: DENIAL

    Shock moves to denial very quickly (usually the 2nd sentence in a posting on a benchmark loss). Comments are usually baseless attacks without any analysis or technical basis. Key words: obviously, clearly...

    Originally posted by Jimmy
    It's well known that Kubuntu provides the worst KDE experience. I'm sorry but anything from Kubuntu is not a fair test.
    Originally posted by BenderRodriguez
    Something's wrong with your setup mate. My Gentoo KD...
    Originally posted by zoomblab
    Something is wrong with these memory results
    Originally posted by kraftman
    This comparison is misleading: (link)
    Originally posted by poofyyoda
    Heh, on my laptop with KDE 4.3 the battery gets about an extra half hour of life [...] compared to Gnome (with compiz)
    Originally posted by <<atomic dude>>
    Seriously flawed.
    Stage 3: DISCREDITATION

    We're geeks, so we look for a technical reason for the loss. Typical postings: 'we lost because of debug symbols', 'they left the default config', obviously the problem is in other component', 'they don't know how to test'. Most don't leave this stage.

    Originally posted by migizi
    You need to rethink these tests. The only true way to know how the DE performs is to do a vanilla install.
    Originally posted by KAMIKAZOW
    Not that everyone already knows that KDE SC components are highly integrated and after startup [...] . In the end it roughly levels out among all of them. [...] I assume that Ubuntu's KDE SC runs the Python-based printer-applet which on its own eat 20 Mb.
    Originally posted by NeoBrain
    I was just wondering whether we can actually trust the results of the KDE 4 memory usage; I'm not sure how polished Ubuntu...
    Originally posted by <<atomic dude>>
    I think, the main problem is the methodology of the measurement.
    Stage 4: ANALYSIS

    Facts are checked, issues and underlying causes are understood.

    Originally posted by Elv13
    KDE so some caching (I don't know what the settings are in Kubuntu, but it's probably enabled).
    Originally posted by energyman
    Don't forget that KDE also provides a lot more features out of the box. Or does XFCE or Gnome have Nepomuk enabled?
    Originally posted by V!NCENT
    Kwin has also more plugins and effects than Compix and what effects are enabled and installed also makes a difference...
    Originally posted by bash
    [...] if GNOME had higher memory usage compared to KDE, most people would have come and said how this proves once and again why GNOME fails for some many reasons.
    Originally posted by mugginz
    Some numbers for your consideration: [...] Clearly Gnome in Alpha 3 has a better memory footprint than KDE no matter whatever KDE is sitting on Ubuntu or Kubuntu. There is considerable fluctuation between boots. This could be due to various reasons.


    Stage 5: ACCEPTANCE

    If you stay the course and make it through the analysis, you ultimately accept the result. The reason for the loss is internalized and implemented upstream.

    Originally posted by <<atomic dude>>
    The sad truth is that many desktop applications are using too much memory these days. KDE uses QT. The memory usage of KDE may be mainly caused by the fact that it uses QT.
    Originally posted by wazyk
    OK, KDE uses more memory than other DEs.
    Originally posted by BlackStar
    Pretty good resulst, I think.
    Originally posted by V!incent
    Results are as expected...And they don't mean anything
    Originally posted by karl
    [...] results are as expected.
    Originally posted by BenderRodriguez
    The whole what eats how much memory is pointless.
    Originally posted by aavci
    I don't think anybody is saying that KDE uses less memory than GNOME or whatever.
    Originally posted by susikala
    KDE is crap, don't use it. These tests show it clearly.
    Originally posted by V!INCENT
    I have 8Gb of RAM. What KDE puts in my RAM I don't care about.
    Originally posted by BenderRodriguez
    No one here argues if KDE uses more memory than Gnome because it's clear I think it uses a bit more.
    Originally posted by SkyHiRider
    I'd say memory is not an issue today.
    Originally posted by damentz
    All you flamers are stupid.
    Originally posted by pfunkman
    RAM is cheap as hell.

    Comment


    • Respekt!

      The k is intentional.

      Comment


      • @yotambien

        So ya got more popcorn on the stove?

        Comment


        • Originally posted by mugginz View Post
          @yotambien

          So ya got more popcorn on the stove?
          Woah, an almost unlimited stock ; )

          By the way, I forgot to mention the Phoronix article from which I ripped the benchmark thing (just in case some newcomer thinks I made it up). Also, it goes without saying that I totally cut and pasted stuff with the sole intention of maximising the comical effect. No attempt whatsoever was made to present an objective summary of the thread and the mentioned authors' posts. I think it's pretty fair, though. : D

          Comment


          • The Meta Stage

            THE NEXT STAGE OF BENCHMARK LOSS

            Stage X: THE META STAGE

            Someone with enough spare time will go through all posts in a forum which already contains a lot of messages. The poster creates some sort of a subjective categorization and goes into great length when fitting each post into the set of those categories.
            Typical postings: 'Stage 1: NAME1', 'Stage 2: NAME2', 'STAGE 3: NAME3', ...


            Originally posted by yotambien View Post
            THE FIVE STAGES OF BENCHMARK LOSS

            Stage 1: SHOCK

            The first reaction when you see an article, scanning down to see your competition wipe the floor with you.
            Typical postings: 'WTF', 'no way', 'you lie'...


            Stage 2: DENIAL

            Shock moves to denial very quickly (usually the 2nd sentence in a posting on a benchmark loss). Comments are usually baseless attacks without any analysis or technical basis. Key words: obviously, clearly...

            Stage 3: DISCREDITATION

            We're geeks, so we look for a technical reason for the loss. Typical postings: 'we lost because of debug symbols', 'they left the default config', obviously the problem is in other component', 'they don't know how to test'. Most don't leave this stage.


            Stage 4: ANALYSIS

            Facts are checked, issues and underlying causes are understood.

            Stage 5: ACCEPTANCE

            If you stay the course and make it through the analysis, you ultimately accept the result. The reason for the loss is internalized and implemented upstream.

            Comment


            • Originally posted by yotambien View Post
              THE FIVE STAGES OF BENCHMARK LOSS

              Stage 1: SHOCK

              The first reaction when you see an article, scanning down to see your competition wipe the floor with you.
              Typical postings: 'WTF', 'no way', 'you lie'...
              Oh yotambien, I think you forgot to wear your glasses before reading the posts.

              Nobody ever said that KDE uses less memory than GNOME. Everybody including me just _stated_ that it uses much less than advertised on the front page.

              Comment


              • Originally posted by yotambien View Post
                By the way, I forgot to mention the Phoronix article from which I ripped the benchmark thing (just in case some newcomer thinks I made it up). Also, it goes without saying that I totally cut and pasted stuff with the sole intention of maximising the comical effect.
                I don't agree with those slides made by Michael Larabel&co. They are implying that Larabel is always right and benchmark critics are always wrong. That's a completely ridiculous implication.

                No attempt whatsoever was made to present an objective summary of the thread and the mentioned authors' posts. I think it's pretty fair, though. : D
                You just confessed that you put those excerpts there without actually understanding them. Nice.

                Comment


                • 141 posts to go

                  This is amusing. Proceed.

                  Comment


                  • @Yotabien:
                    You FAIL.

                    Too bad there's a 1min edit tinme limit, because I would have demanded you took away my quotes.

                    Nice BS theory of stages, but don't quote me in the WRONG ORDER!

                    I said from the beginning the resulst were expected. Kwin was compared to Compiz which has NOTHING to do with Gnome.

                    Then you quoted that I said "I don't care for the RAM usage because I have 8GB", but you failed to miss the point COMPLETELY.

                    *Oh BTW... I guess this is my denial stage... yeah you fail terribly*

                    My point was that one should look at what RAM is executed. The entire idea of sucking up RAM doesn't mean using/changed it and thus means absolutely zero in speed/performance, energy consumption and CPU load. I'm saying it again; you failed.

                    Furthermore KDE has shown in the past (with KDE 3.5.x) that when it's feature stripped to the low point of Gnome it has less RAM, less CPU load and thus less battery consumption and is faster. What that means is that Gnome should get first to the point of KDE 4.4.x's features before even being allowed the right to speak out/spoken for.

                    But like BlackStar said: this is amusing. Although correction half-truths and half-quotes tires me, it's endless satisfaction

                    Comment


                    • 139 posts to go

                      My point was that one should look at what RAM is executed.
                      KDE is now executing RAM?! Bug number? Are virtual machines safe, or are they also affected? I always knew KDE was buggy, but I can't believe they shipped 4.4.1 with such grave bugs!

                      Comment


                      • btw, how was memory consumption measured? ps aux? free? a different way to completely fail?

                        Comment


                        • Originally posted by yotambien View Post
                          THE FIVE STAGES OF BENCHMARK LOSS

                          Stage 1: SHOCK

                          The first reaction when you see an article, scanning down to see your competition wipe the floor with you.
                          Typical postings: 'WTF', 'no way', 'you lie'...


                          ..clipped...
                          Miguel, do you think people don't see through your agenda?

                          Comment


                          • Originally posted by ender2070 View Post
                            Miguel, do you think people don't see through your agenda?
                            Sarcasm..?

                            Comment


                            • Originally posted by << ⚛ >> View Post
                              I don't agree with those slides made by Michael Larabel&co. They are implying that Larabel is always right and benchmark critics are always wrong. That's a completely ridiculous implication.
                              Completely wrong. (I was the primary author, and coined the concept - so this is one of the times that I can actually use "completely".

                              The point is that the benchmarks are just what they are. Benchmarks.
                              The numbers are just what they are, numbers.

                              I believe that within this thread *NOONE* has gone through and reproduced *EXACTLY* what Michael did for the article. mugginz came close, and guess what, he was in a similar order.

                              The majority of this thread is about people arguing their side of the coin, or saying that their figures don't match up. As mentioned in the presentation, and mentioned elsewhere.

                              Feel free to standup and do the following

                              1) Reproduce the test and the results
                              2) Make suggestions for improvements or clarifications
                              3) Present a similar set of results with the modified methodology.

                              But I don't think I have seen anyone get past 1). No one has voting rights for 2 or 3 unless you do 1.

                              You just confessed that you put those excerpts there without actually understanding them. Nice.
                              Quotes are just that. Out of context they can be used anyway. Fortuntately, with digging the context can be found for all the quotes.

                              Regards,

                              Matthew

                              Comment


                              • Got bored and wrote some automated bench scripts.

                                First lot off the press.



                                Bench procedure: (Chart 1)
                                • Clean install from standard ISO of Kubuntu Karmic 9.10 (no updates)
                                • Apply the bench script and video file to the install just after installation procedure and before machine is booted.
                                  Note: Numbers for Run 1 represents the first time that image has ever been booted.
                                • The machine is booted. Bench script is auto-run on each boot.
                                • Bench runs and when finished auto reboots machine, etc.
                                • Once 5 loops of Boot-Bench-Reboot occur, metrics are stored to another machine.

                                After the fifth run of the first group, the bench script stops and all updates are applied. (Standard repos)

                                Bench procedure: (Chart 2)
                                • The machine is booted. Bench script is auto-run on each boot.
                                  Note: Numbers for Run 1 represents the first time that image has been booted since updates were applied.
                                • Bench runs and when finished auto reboots machine, etc.
                                • Once 5 loops of Boot-Bench-Reboot occur, metrics are stored to another machine.

                                The contents of the benchmark are as follows:
                                • Autostart on boot recording RAM usage once every second for 20 seconds.
                                • Launch Konsole and record RAM usage once every second for 20 seconds.
                                • Quit Konsole via dbus and record RAM once usage every second for 20 seconds.
                                • Launch Dragon Player with ogg theora video and record RAM usage once every second for 20 seconds.
                                  (Dragon Player is the default video playback program for Kubuntu 9.10)
                                • Quit Dragon Player via dbus control and record RAM usage every second for once 20 seconds.
                                • Reboot the machine.

                                Interestingly the machine never saw more than 267M used if you dont include the first run after installation or updates (Run 1 for both Charts)

                                These numbers are for x64 installs and KDE 4.3

                                The next run will be for Ubuntu 10.04 Gnome x64.

                                Comment

                                Working...
                                X