2017: Should Linux Benchmarking Still Be Mostly Done With Ubuntu?

[shmerl]

I disagree that Ubuntu is the most popular desktop distro. It's not anymore.

[mike44]

ok Mint as from distrowatch. But that's based on Ubuntu.

[Marc Driftmeyer]

I've run Debian Sid for the past 17 years. Games should be tested on that level of stability. Stable is too conservative.

[kaprikawn]

I have an issue with the whole 'Ubuntu is most popular so is the obvious choice' thing. It may be popular amongst the entire population of Linux desktop users, but the audience we're talking about is Phoronix readers. Is Ubuntu the most popular amongst Phoronix readers? I would content not, admittedly without any solid evidence.

Phoronix readers tend to be high information users. And so we know all the bull that Canonical pulls. So contrary to the assertion that Ubuntu is popular, I would say that amongst Phoronix readers Ubuntu is decidedly unpopular.

I'm not arguing that that Ubuntu shouldn't be the choice, I'm just picking issue with one of the major arguments given. I personally feel Fedora would be a good choice these days, but I have no major issue with Ubuntu being used (until it moves further away from the standard as mentioned in the article).

[aufkrawall]

I'd like to see benchmarks with up to date components, and this is not the case with frozen releases like Ubuntu.
Thus I'd be be most interested in Arch/Tumbleweed testing platform, and for filesystem ext4 with active trim on SSDs, like it is usually recommended.
Also CPU governor should be set to a setting which works best for AMD/Intel (still powersaving, but speed considered more important). Examining this would make a nice article, btw.

Oh, and like I said before: When there are benchmarks done which stress the filesystem on different distributions, it would be nice to test apples to apples (aka ext4 active trim vs. ext4 active trim). If someone wants to find out how filesystems compare, they should be tested with the same distribution (which should be again Arch or Tumbleweed for having everything up to date).

[Darknation]

Personally haven't used Ubuntu in years for my desktop OS. But it is still the most popular distro in one form or another.

I personally use Antergos and I can't fault it. Although I want to build my own Arch at some point.

Use Arch Anywhere, it helps to simplify setting up an Arch system.

[sunweb]

Kubuntu is apparently a pretty bad KDE distro because they ship outdated Qt versions usually.

I personally would like to see Fedora / openSUSE / Arch if Ubuntu is no longer the main distro. Doesn't really matter which one.

Frankly going by your logic Manjaro and Arch are also bad choices. They shipped Qt 5.8, sure its new but it brought alot of bugs to KDE that is considered stable with Qt 5.7. So in reality all this bullshit must disappear only with standalone packages(flatpack or snap) when users will run programs as they meant to be by developers of the software and don't have additional bugs because of too slow/fast distro maintainers.

Basically, before standalone packages come into play i think its better to setup Gentoo for tests.

[dungeon]

Whatever, on average if drivers and apps works OK - any Linux distro is OK

[dimko]

As long as Ubuntu is not using Unity - it's fine.
No Unity please, only Xorg.

[juno]

Canonical is not in any way relevant. The default linux desktop is developed at redhat. Ubuntu does only adopt - months later.
Also, people use rolling release distributions. They also don't want to use damn broken SteamOS and Steam runtime with tons of ancient versions of shared libs in their ~. Even Valve acknowledged that now.

I vote for a rolling release distribution as default. Steam games should be tested w/o steam runtime. What's the point when testing e.g. the newest driver releases or even -git snapshots with an ancient system? Once in a while there should also be a benchmarking session vs. the official latest release of SteamOS