What Linux Distribution Should Be Benchmarked The Most?

[Sidicas]

I would prefer Debian with Mesa compiled from upstream source..
Second to that, I'd say Mint. It seems more popular than Ubuntu these days.

[jmartins]

+1 Manjaro Linux and Arch Linux

Chakra is good too but only KDE

[0xBADCODE]

Ubuntu flavors of course. Maybe mint.

Primary testing OS should be Ubuntu. Maybe some Mint as well. No point to pay too much attention to things which are not really popular. While every fan would insist their linux distro is the best, this test would be fairly useless for most users around. Because only few gurus are actually using other distros. After all, all those arch, gentoo and so on are not user friendly and it takes some "rocket science" to set them up and keep things running withouh issues. This dooms them to be unpopular in general public.

However I think it's okay to test different distros from time to time. Say, Fedora is a RedHat testbed and redhat does impressive amount of work on diffrent subsystems. So it could be worth to test it sometimes. Then there is OpenSUSE. Probably another adequate candidate as desktop OS.

[jijitus]

Create a poll!

I confess I've read the first four pages of the thread only.

The question is not wrong, but all people will answer according to their own preferences, and some are even evangelizing their Chosen One. So the question is a flamewar-inducing one.

Michael: why don't you create a poll? Something small where every Phoronix user can check all the distros they like - yes you might need to include the "stable" and "latest" of a few distros to keep everyone happy.

[simonandric5]

1. bodhi linux
+
--------------
2. lubuntu
3. kubuntu
------------



simon

+ my favourite


kubuntu PROJECT NEON


s.

[bigears]

Based on workloads

The majority of benchmarks on Phoronix is compiled, let it be filesystems to drivers. So break it into 3 categories and in each category choose a distro. The distro's choice should be based on flexibility, that is, how many hardware platforms can it be installed on. For example, Ubuntu is going to be installed on workstations, phones, TV, etc but Mint will probably be not, no matter what direction Canonical has taken. Similarly, Gentoo has a large hardware footprint due it been source based, hence preferred by the performance crowd, likewise servers are based on stability and security.

Category	Distro	Why	Vote
Popular	Ubuntu/Mint	This will please the workstation crowd and the less power users	Ubuntu
Performance	Gentoo/Sabayon	This will please the workstation power/small server sysadmin users	Gentoo
Server	Centos/Gentoo	This will please sysadmins and security users	Centos

Benchmarks should be based on workloads not popularity. After all this is not the 'X factor'. It is not hard.

[GreatEmerald]

Michael: why don't you create a poll? Something small where every Phoronix user can check all the distros they like - yes you might need to include the "stable" and "latest" of a few distros to keep everyone happy.

You're clearly missing something.

[korrode]

Arch / Manjaro
If Arch isn't suitable for some reason; openSUSE.

[ibrahimsuprapto]

Dear Phoronix, This Distro is a must

I think This Distro ...

1. Ubuntu LTS because it's for business and desktop user

Optional :

2. Centos/Redhat because it's used for Enterprise user
3. Arch Linux, Simple, minimal and Rolling release Distribution
4. Distro Major : Slackware<tgz based>, Redhat<rpm based>, Debian<deb based>

[teotwawki]

Ubuntu

I'd say Ubuntu for the following reasons of basic logic:

As you've already said, it's very popular and has LTS stability for comparability over time.

From a server point of view, it spans casual & serious users because formal support is available but optional.
You mentioned that most Linuxes are pretty close on performance under the hood, so benchmarks looking into filesystems or new CPUs etc. should be just as insightful for CentOS & other server users.

When then main Ubuntu goes to Mir, there will doubtless be forks such as there are with Kubuntu etc.
These derivatives from a common base allow comparing the effect of the different GUI layers.
It also means that benchmarks unrelated to the GUI layers are just as informative for users of the derivatives.

Finally, better the devil you know. You are the one doing all the work so if there are aspects of Ubuntu that make your current workflow easier then any change is likely to reduce the amount of quality articles we get to read.