Ubuntu 12.04 Desktops Impact Performance, Power Consumption

[phoronix]

Phoronix: Ubuntu 12.04 Desktops Impact Performance, Power Consumption

How does the choice of desktop environments you make for Ubuntu 12.04 LTS impact your system's performance and power consumption? Here's the latest round of benchmarking from the various Ubuntu 12.04 desktop environment choices -- Unity, Unity 2D, GNOME Shell, KDE, Xfce, LXDE, and Openbox -- when running them on three different laptops.

http://www.phoronix.com/vr.php?view=17241

[cl333r]

For Christ's sake please use different colors. I have to try really hard to detect which is which: Xfce vs Gnome Shell vs LXDE:

http://openbenchmarking.org/embed.ph...ha=44d0b03&p=2

[mdias]

I agree with the previous comment.

4 shades of blue, no orange, no pink, no grey...

[Michael]

Patches welcome to improve the color selecting algorithm... I guess there's limitations to the latest algorithm that was submitted by a Phoronix reader. Those interested can look inside pts-core/objects/pts_Graph/pts_Graph.php.

[goretrade]

I'm just curious about it, since I never saw tests using really tiny wm or without it

[russofris]

I'm just curious about it, since I never saw tests using really tiny wm or without it

I'm almost as curious to see a "no wm" reference score, though I do not know if most of the apps will run without one.

[krazy]

Rather than just noting the capability, it would have been useful to benchmark KDE with unredirected fullscreen windows. Otherwise the numbers aren't useful - it's apples-to-oranges..

[cl333r]

Patches welcome to improve the color selecting algorithm... I guess there's limitations to the latest algorithm that was submitted by a Phoronix reader. Those interested can look inside pts-core/objects/pts_Graph/pts_Graph.php.

Why don't you use Highcharts instead of using your own solution? afaik it's free and does client-side rendering, it interacts with the mouse, looks professional and is much more handy, works in all browsers etc.

[Michael]

Why don't you use Highcharts instead of using your own solution? afaik it's free and does client-side rendering, it interacts with the mouse, looks professional and is much more handy, works in all browsers etc.

I've commented on the many reasons before in other threads.

[curaga]

I'm almost as curious to see a "no wm" reference score, though I do not know if most of the apps will run without one.

It will improve input latency, and most games should run. As to output, I'd like to know the exact numbers too. +1