Announcement

Collapse
No announcement yet.

100+ Benchmarks Between Clear Linux vs. Arch-Based Manjaro Linux - Summer 2019 Tests

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • make_adobe_on_Linux!
    replied
    Sweet benchmarks. I would love to see a representative from each distro (or something) be in charge of configuration. And Gentoo would be especially interesting to see - I've not seen in benchmarked for quite a while. It is supposedly "not needed anymore since our hardware is so fast." But a posh bunch of Phoronix graphs on EPYC optimized for Gentoo would be interesting to see. Also would be interesting to see how much worse the zen kernel performs on such benchmarks.

    Leave a comment:


  • Michael
    replied
    Originally posted by jawz101 View Post
    I like the new graphic. It's easy to see who wins. Maybe conditional coloring to highlight significance. If people don't like the long vertical, perhaps chunking it out into a few dashboard widgets but you could still see it on one page. Maybe chunk it out on classifications of tests and then for each section denote an overall winner, and have an overall winner. Throw some statistics in how it weighs things or whatever.
    Thanks for feedback.

    There is 'overall winner' denoted in other graphs like the geometric mean graph. As for dashboard widgets or showing more stuff embedded in a single graphic, that isn't done since I am maintaining the same code path to also be able to generate graphs for PNG (in cases of old browsers still or other odd scenarios) as well as for being able to generate the graphs for embedding into PDF files, etc.Rather than maintaining multiple code paths due to being short on time / resources as it is, that is unfortunately why more graphs aren't 'interactive' right now.

    Leave a comment:


  • jawz101
    replied
    I like the new graphic. It's easy to see who wins. Maybe conditional coloring to highlight significance. If people don't like the long vertical, perhaps chunking it out into a few dashboard widgets but you could still see it on one page. Maybe chunk it out on classifications of tests and then for each section denote an overall winner, and have an overall winner. Throw some statistics in how it weighs things or whatever.

    Leave a comment:


  • rabcor
    replied
    Originally posted by petterk View Post

    It's not Arch.
    Wow, the perfect argument. I guess nobody should ever use anything else than arch should they?

    (also you're kinda wrong, it more or less is arch)

    Leave a comment:


  • petterk
    replied
    Originally posted by GdeR View Post

    People like you made me want to stay away from Arch. The Arch community is full of childish bigoted zealots, hidden behind a nice Wiki.
    Manjaro is not Arch so while you were offended at the Arch forums for being told you're not using Arch, you were simply being told the truth.

    Leave a comment:


  • petterk
    replied
    Originally posted by rabcor View Post

    Why not manjaro?
    It's not Arch.

    Leave a comment:


  • petterk
    replied
    Originally posted by kaprikawn View Post

    If what you get on Manjaro is compiled without optimizations and what you get on Clear Linux is compiled with -O3 then the results are valid. What would be the point in installing a distro, then building everything from source? If you go down that road, everything Michael does is pointless unless it's on an LFS system.
    Ever heard of Gentoo..?

    Leave a comment:


  • partizann
    replied
    Also quite like the new 1on1 graph for such comparisons.

    Leave a comment:


  • NotMine999
    replied
    I like the new graph format. Very easy to read in 1 screen, or scrolling in 1 screen.

    I will echo a comment from another poster: I would like to see most divergent/significant results at the top with the "no difference" results at the bottom.

    As of OS choices, as long as the OS choices are tested using the same tests, then I think the results ARE VALID for comparing OS.

    How an OS chooses to recompile source or choose options for a package should be reflected in these tests. Some distros tend towards conservative options while others "push the limits", and I find that an interesting comparison.

    Leave a comment:


  • lsatenstein
    replied
    The new graph is counter intuitive. I prefer one of two options.
    Two single graphs, with same measured results pointing to the right.
    Or the conventional double-bar graph with one bar Manjaro, the other Clear-Linux.

    Next month, can the test be repeated using AMD's cpu

    Leave a comment:

Working...
X