Benchmarking: Ubuntu v/s Fedora

[J]

On Tue, Jun 19, 2012 at 9:45 AM, Steve Flynn <anothermindbomb at gmail.com> wrote:
> On 19 June 2012 14:36, Steve Flynn <anothermindbomb at gmail.com> wrote:
>> On 19 June 2012 14:30, H.S.Rai <hardeep.rai at gmail.com> wrote:
>>> Does information at
>>>
>>> http://openbenchmarking.org/result/1206074-SU-1206062SU09
>>>
>>> is authentic?
>>
>> Yep - Phoronix has been benchmarking Linux distributions and kernels
>> for quite some time. No reason to doubt the reliability of the
>> numbers.
>
> Gmail interface sends when you hit Tab + enter... always catches me out.
>
> I was about to say that these particular results are from similar
> spec'd machines but are completely different version of the kernel,
> compilers, display servers and so forth, so making a comparison
> between them is relatively pointless.

I'd say that is an entirely valid comparison.  It's comparing the
latest current releases from each distribution, Ubuntu 12.04 and
Fedora 17.  Would you say that comparing Windows 7 to Ubuntu 12.04
would be invalid because they use completely different display
drivers, kernels and so forth?

The kernel version, patch level, compilers, library versions and so on
are chosen by the respective distributions for their release, so the
differences are intentional, not accidental.  The only point that I'd
say makes this somewhat unfair is that Fedora 17 is a playground for
RHEL, where Ubuntu 12.04 is more akin to RHEL itself.  So a MORE
proper comparison would be the current RHEL to 12.04, but comparing to
the current Fedora, OpenSuSE, Slackware, Debian, etc is still valid.

It would be different if they were comparing, say, Lucid to Beefy Miracle.