And yet my own personal experience was that Arch couldn't play 720p x264 video on my Linux machine, but Slackware runs them all just fine.

After some research I can confirm that slackware was the only distribution running a 3.x kernel at that time.
(Not sure about arch). What I think being the biggest mistake of this test is testing a beta distribution along with stable distributions, they could take 13.37 and get the same result for every distribution since imho there is no such thing as a "benchmark distribution". They could test different kernels on the same distribution and get the same result they got in this "distribution test"

Just don't trust phoronix, period.

Regards

I don't trust phoronix at all, I don't go there anymore because of the ridiculous benchmarks they do. They're not even benchmarks, they are nonsense drivel.

I will likely do a proper benchmark of Arch and Slackware 14. However, it's not that useful because it isn't the distro that is fast or slow, it's the software and the kernel. Still, I would like to see if there is a huge difference between the two, even tho there is no chance that I will switch to Arch.

Until recently, my neighbor here in my little South French village drove a Suzuki 1100 GSX-R motorbike. He's been bragging quite a lot about the incredible acceleration, top speed, and so on. Yet I saw him spend most of the time repairing his bike, because there was always something that had to be fixed. At the same time, my own 750cc BMW motorbike - the same model than the one from the French police force - doesn't have much to brag about. It's reasonably fast, reasonably comfortable for me and my girlfriend, I drive it all year in summer and in winter, and it takes me everywhere across the country, and sometimes even to Switzerland, to Italy, to Austria or Hungary. It just works, without much fuzz.

A few weeks ago, my neighbor was sick of his Suzuki and traded it for the exact same 750cc BMW that I have.

For motorbikes as for computer systems, the real benchmark test is not "How fast does it go?" but "How far can I go with it?"

@kikinovak, +1

According to the link the OP posted:
A dubious benchmark was done on different hardware with a 64GD SSD that noone is using for real work, comparing an unreleased version of Slackware, a "harder to do" default configuration of Arch and Fedora, CentOS and Ubuntu with "stock/default configuration as much as possible" "to largely appease the Phoronix readers"...

Which actually can be read "I have to deliver that article for Phoronix, but I have no material... what can I do... wait! I have an idea! ..."

That was a beta version of Slackware 14 ... that doesn't count, it is unfair, i hope they can do the comparison once the stable release of Slackware 14 appears ...

I recently cancelled my RSS subscription for Phoronix because of the serious decline of the articles' quality. I don't trust benchmarks from them at all, you can see the quality from that article alone in the listing of the configuration (Arch Linux with no desktop, X and OpenGL listed, Disk Details only listed for Ubuntu, Processor details missing for Slackware) and the diagrams (Arch missing in the BYTE benchmark).
This benchmark is simply useless.

I have used every version of Ubuntu since the inaugural 4.10. I have been using Slackware since version 11.0.
In my experience, Slackware has always run faster and used fewer resources than Ubuntu, especially when using XFCE on Slackware.

Note that Phoronix has documented that Ubuntu has been getting slower and more bloated ever since Ubuntu 8.10:
http://www.phoronix.com/scan.php?pag...nch_2008&num=1
And more recently since Unity came along:
http://www.phoronix.com/scan.php?pag...desktops&num=1
http://www.phoronix.com/scan.php?pag...unity_64&num=1
This is why I have long ago switched to the light and fast Lubuntu with the LXDE desktop instead of Ubuntu with Unity.

Benchmarks only matter when your distro/os comes out on top.

cheers,

Lol, comparing a BETA version of Slackware against STABLE versions is so FAIL.

Curiously, the article has a point at least in this regard.

• There's no stable version of Arch. It's perpetual Beta.
• Ubuntu draws more than 90% of its packages from Debian testing and unstable repos.
• Fedora is Red Hat's self-proclaimed test bed for new technologies.

Which leaves only CentOS as the actually stable and rock-solid contender.

I saw the word "phoronix" and didn't bother clicking - sensationalist tech press nonsense... their very existence depends on publishing crap like that...

Slackware (with KDE 4.6.5) is the fastest, most stable GNU/Linux distribution on my quite humble system. Which is the "benchmark" that matters most to me.

True.
Which is 10% more stable than Slackware 14 BETA? :)
Alright, did not know that.

Ok, you win. Now I have to go back to work.

I checked out the chart, looks to me the systems used have some minor differences (memory config different, GPU ...). But what I am really curious about are the compiler details. I am far from a 'benchmark person', but wouldn't you want to run the same statically linked binary across all Linux systems ?

John