OpenBenchmarking.org: Ubuntu Prepares To Be Overtaken By Arch Linux

[dh04000]

LOL . Arch folks don't usually compile stuff though, unless it's from AUR.

All from AUR or get out of Arch!

[Creak]

The reason Arch popularity has skyrocketed is the Evolution arch installer.
Makes it really simple to get arch up and running for us non "gifted" people.

Do you have a link for that "evolution" stuff? Installing Arch was fun... once.
Of course I didn't write all the packages I installed to have a nice GNOME installation with all the services working properly.

Thanks!

Edit> Think I found it: http://www.evolutionlinux.com/

[Creak]

This "only for bearded hackers" thing from Arch reminds me of the old Debian installer without GUI.
In the end, GUI installer will officially happen on Arch, just wait... a long time, but wait

[alaviss]

Do you have a link for that "evolution" stuff? Installing Arch was fun... once.
Of course I didn't write all the packages I installed to have a nice GNOME installation with all the services working properly.

Thanks!

Edit> Think I found it: http://www.evolutionlinux.com/

Beware that using it means that you can't get help on Arch BBS or mostly others Arch Community, they won't support any Arch-based distributions even if that distro get files from Arch Linux official mirrors

[Michael]

If the testsuite behaves on other fedora installations as on my, I understand that its % is so low: Out of curiosity I tried to run some of the test - already at downloading/installing the missing pakets, it crashed/went into some kind of endless loop that created as many processes as possible. Also the system detection is buggy: System info yields with my system: PROCESSOR = Intel Core i7-3930K @ 5.70GHz (12 Cores) - and no, my system is actually not running at 5.7 GHz.

I found the idea of such a test suite very good, even when it is not running on all systems/distributions (I am not curious enough/have that much spare time to track that bugs back). Michael does a good job with this, but one should be careful when drawing general conclusion as "this/that distri is the most popular for conduction test", when maybe the reason is just that it is not working out of the box with certain distributions.

Strange, I routinely use PTS on Fedora and never run into problems... The only problems most people run into are when the HTTP mirrors of files disappear and then eventually I update them when notified of downed mirrors. What test(s) were you trying?

Can you try the latest PTS code to see if it still reports 5.7GHz? I have not encountered such bogus frequency detection before, so likely either your CPUfreq/P-State driver or /proc/cpuinfo is somehow supplying the incorrect data.

[johnc]

This is my first time hearing of /etc/os-release. Was there some valid reason to move away from /etc/lsb-release (massive security hole in a text file, original author was a murderer, etc.) or is this just another one of the all-too-common examples of some Linux dweeb just wanting to do things differently?

[Kemosabe]

This is my first time hearing of /etc/os-release. Was there some valid reason to move away from /etc/lsb-release (massive security hole in a text file, original author was a murderer, etc.) or is this just another one of the all-too-common examples of some Linux dweeb just wanting to do things differently?

Just guessing but the following scenario sounds plausible: The original murderer-killer-IS-gangbang-criminal-drug-abuser-author of lsb-release specified lsb-release textfile in an insufficient manner. My arch provides lsb-release and os-release (old installation whatever dont know, dont care) and there are significant differences with latter provding more information.

[tomtomme]

wrlinux

what is wrlinux???

[johnc]

Just guessing but the following scenario sounds plausible: The original murderer-killer-IS-gangbang-criminal-drug-abuser-author of lsb-release specified lsb-release textfile in an insufficient manner. My arch provides lsb-release and os-release (old installation whatever dont know, dont care) and there are significant differences with latter provding more information.

That's true. I could see how it would be an overwhelming difficulty to add an additional line or two to the existing lsb-release text file. It would clearly be easier to just create a whole new file and move everyone over to that.

[blackout23]

This is my first time hearing of /etc/os-release. Was there some valid reason to move away from /etc/lsb-release (massive security hole in a text file, original author was a murderer, etc.) or is this just another one of the all-too-common examples of some Linux dweeb just wanting to do things differently?

http://www.freedesktop.org/software/systemd/man/os-release.html

The /etc/os-release symlinnk is also only for legacy compatability. Vendor defaults should go into /usr/lib, user overwrites should go into /etc this way you do factory resets by nuking /etc and boot with only /usr populated which is esentially the entire distro.

Blacklisting for the nouveau module when you install the nvidia driver also happens in /usr/lib/modprobe.d/nvidia.conf not /etc/modprobe.d