Vote On A Distribution For Linux Benchmarking

[Michael]

In continuation of the discussion started yesterday about which Linux distribution to benchmark on Phoronix in place (or complement) Mir-based Ubuntu for graphics testing, etc... Here's a poll. Pick your preferred Linux distribution to see benchmarks of on Phoronix.

[TheBlackCat]

I personally don't get the point of the poll. Shouldn't the distro chosen be the one most technically suited to testing?

[test testing]

Where is Linux Deepin in this poll?

[przemoli]

OpenSuSE.
They support all the DE's out there (that can be "easily" packaged into OpenSuSE) so less likely occurence of DE being configured in less then optimal way.
There is corporate equivalent (so work on enchancing those test would help other phoronix clients?).
OpenSuSE is supported by AMD and Nvidia (through SLED but ..), so bugs specific to it should get a bit more attention.
EXTENSIVE 3rd party repos. (More stuff can be downloaded via binary packages)
OpenSuSE Build Service. (Could be used to spin off, custom linux distro for PTS , but its not OpenSuSE specific..)
Kernel git repo with .specs files so one can get daily builds of supported kernel.
Its green colored. (Though others may not share my favourite color. Shame on them :P )

[wizard69]

I voted for Fedora.

I never really understood the fascination with Ubuntu. I alway saw it as a crap botch job of a distro.

Beyond that does it make a big difference. I mean really if you compile a new kernel for testing or new graphics drivers aren't you in the custom distro zone. In fact with all the testing you do I would imagine you wold have to do a lot of clean installs to keep the data rational.

[grenadecx]

I personally don't get the point of the poll. Shouldn't the distro chosen be the one most technically suited to testing?

Only problem is, all are suited. The only reason why this poll exist in the first place is people complaining about the benchmarks that was taken using Ubuntu.

[chrisb]

I personally don't get the point of the poll. Shouldn't the distro chosen be the one most technically suited to testing?

Not necessarily. People might also base their choice on:

• most Likes on Facebook
• highest score on DistroWatch
• most profitable parent company
• most corporate subscriptions
• most widely used as a base/reference
• most pre-install hardware sold
• most supported by third party closed software
• what they personally run

I currently run Ubuntu LTS, Debian Stable, and Raspbian; I actually voted for Debian Sid because this is ultimately the base of all of those, and it is the Linux "bleeding edge" - it has the software that ultimately ends up in every other distribution, so benchmarking it provides information that is applicable to every distribution, including the RPM based ones. Though I was tempted to vote Ubuntu, because Ubuntu is 4x more popular (Facebook Like metric) than Debian, so reaches a wider audience, and in many ways has potential to reach a much, much wider audience if Ubuntu Touch is successful - which is admittedly a big *if*, but if it does happen that Ubuntu start shipping the software of, say, 50 million phones and tablets a year, in addition to the 18 million desktop pre-installs that they expect this year, then the potential is much greater than the Debian project alone. But the underlying software is still mostly Debian, and as far as Mir/Unity go - you never know - it would only take one developer to package them for Debian...

[Pallidus]

what a waste of time this is...

Hey phoronix how about 3 way benchmarks between os x 10.9, win 8.1, ubuntu 13.10 +mesa?

computational benchmarks with h264 encoding/decoding and stuff like that, what really matters.

I swear the singal to noise ratio in here is getting worse, no one cares about sabagayonarchmangay linux

it has been shown time and time again there isn't any major difference in performance between same kernel distros

if people don't like composite or accelerated DE simply test using xfce or lxde and then everyone will be happy

[CrvenaZvezda]

I personally don't get the point of the poll. Shouldn't the distro chosen be the one most technically suited to testing?

My thoughts exactly. Thus I voted for openSUSE.

[rvalles]

Second round

I expect a second round, with a reduced number of options.