"Open-source is bad!"

[BCEmporium]

OK, to give you more data, as I set 3 machines here, all to the same outcome on performance (SuSE was just set in the Sempron LE):

Atom D525 4 Gb DDR3 SO-DIMM - Ubuntu 11.04 64bits - Gnome desktop;
Zacate E350 4 Gb DDR3 DIMM - Ubuntu 11.04 64 bits - Gnome desktop  and KDE (Kubuntu)
Sempron LE-1250 4 Gb DDR2 DIMM - Ubuntu 11.04 64 bits - Gnome desktop

From the applications one can benchmark - existing for both Windows and Linux that is -, like VLC, Aptana or Libre/Open Office, Linux under performs hardly on all of them.

With an i5, i7... you probably won't notice differences because those machines have power to spare to out perform you on all levels and OS's. At low end where resources are ratiocinated you can easily see the low performance of Linux when compared to Windows.

[1715633675]

1715633675

[1715633675]

1715633675

[jgraham]

OK, to give you more data,

Hey isn't this "thinking inside the box"?  Don't you have a problem with non-linear logic?  

Atom D525 4 Gb DDR3 SO-DIMM - Ubuntu 11.04 64bits - Gnome desktop;
Zacate E350 4 Gb DDR3 DIMM - Ubuntu 11.04 64 bits - Gnome desktop  and KDE (Kubuntu)
Sempron LE-1250 4 Gb DDR2 DIMM - Ubuntu 11.04 64 bits - Gnome desktop

From the applications one can benchmark - existing for both Windows and Linux that is -, like VLC, Aptana or Libre/Open Office, Linux under performs hardly on all of them.

What benchmarks were you running, how many runs did you do, what were the results? (The word "hardly" shouldn't be there - "handily" might be the word you are thinking of but idiomatically it's rarely used with "under perform" only "out perform".)   Not to mention with the exception of Open Office aren't these poor choices for benchmarks? 

With an i5, i7... you probably won't notice differences because those machines have power to spare to out perform you on all levels and OS's. At low end where resources are ratiocinated you can easily see the low performance of Linux when compared to Windows.

Ok, first of all congratulations on using the term "ratiocinated" and further congratulations on using it entirely incorrectly and further fanfare for not realizing it.   I think the word you're grasping at is "rationed".
 
Anyway, so now you're saying that Linux only under performs on slow hardware?  Funny how your thesis just keeps changing...not because the data is changing...but rather because you seem to have used misleading terms.

[BCEmporium]

English isn't my native language and I don't quite give a fuck for your corrections as the message seamed to pass through anyway, it under performs on ALL flavors, you just won't be able to notice such on high end machines, unless you come up with tools to show you that CS is running at 100 fps here and 200 there, whereas your eyes can't see any difference above ~25 fps, so big deal to have 200 fps if for usability it's the same as 50 or 100.

[jgraham]

English isn't my native language and I don't quite give a fuck for your corrections as the message seamed to pass through anyway

Pass through what? ("through" usually requires an object) .  Normally I'm in favor of anything that can be understood but you really push the border there.

, it under performs on ALL flavors, you just won't be able to notice such on high end machines, unless you come up with tools to show you that CS is running at 100 fps here and 200 there, whereas your eyes can't see any difference above ~25 fps, so big deal to have 200 fps if for usability it's the same as 50 or 100.

So by FPS I assume you're talking about VLC now (but you might be talking about monkeys...who knows?).   The problem with this benchmark on modern hardware is that it doesn't have much to do with CPU or the OS for that matter.   Most of the rendering, in most cases is taken care of by the GPU.  All you would be comparing is the difference between VAAPI/VDPAU against DxVA which isn't really telling you very much.   For example my Z-series Atom had much better video under Linux than Windows 7 - both in Gaming and in VLC.  Why?  Because the GMA500 was poorly supported.  On Windows 7 it wouldn't be recognized as a OpenGL provider (I hacked the drivers to get it working but it never was very stable) the support in Linux was better (not great, just better).  In fact I used it to drive a 1080p display with multiple subtitles.  Which never, ever worked under Win7.

The point is that as a benchmark it doesn't say much about Linux generally (even though in this case Linux "won") because it was really measuring Intel's decision to disable OpenGL under Win7 and their commitment to produce good drivers (which was probably tied to the fact that they didn't make the hardware - it was a PowerVr GPU re-branded by Intel)

[BCEmporium]

Again you take everything strictly... 

"Passed trough your head"... and, man, that sounds like a hard to do task!

I wasn't comparing FPS's, I was stating that on high end hardware it will be always - or almost always - above the user's human capacities. So it doesn't matter if you use Win/Linux/Solaris/BSD/MacOS... it will flow smoothly enough. Unless you've some benchmarking numbers in the corner of the screen - means; if you're a freak gamer - you wouldn't be noticing any difference within that example, even if CS ran at 100 fps under OS X and at 200 fps at OS Y.

[jgraham]

Again you take everything strictly...  

"Passed trough your head"... and, man, that sounds like a hard to do task!

See I would have assumed "passed though BCEmporium's ass".  I can't comment on the difficulty.

I wasn't comparing FPS's, I was stating that on high end hardware it will be always - or almost always - above the user's human capacities.

You sure mentioned them.  The point though, if you read what I posted.  Is that these values rarely have anything to do with "high-end hardware" or OS's.   For lots and lots of benchmarks FPS is just going to test the video hardware.   Didn't you notice my example had nothing to do with high-end hardware?  Apparently not.