Announcement

Collapse
No announcement yet.

Linux On IBM Z "s390" To See ~11% Higher Syscall Entry Performance On Linux 6.8

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

  • #11
    Originally posted by rene View Post
    That being said, it should be pretty obvious that Phoronix does not meet professional journalist reporting standards, and is usually expressing their personal opinion in the articles and giving them a rather positive or negative coloring accordingly.
    In all fairness this type of content does lend itself to the author expressing an opinion because they are reviews, not neutral reporting.

    A review by it's very nature is what the reviewer thinks about the thing he is reviewing.

    If you want to see bias that resulted in poorly done testing, look back at the old HardOCP reviews where that guy was such an Intel shill that Intel eventually offered him a Director of Marketing position and he shut down the site to take the job.

    I used to love HardOCP and his reviews but Jesus Christ could that guy slant a review in Intel's favor.

    Comment


    • #12
      Originally posted by sophisticles View Post

      In all fairness this type of content does lend itself to the author expressing an opinion because they are reviews, not neutral reporting.

      A review by it's very nature is what the reviewer thinks about the thing he is reviewing.

      If you want to see bias that resulted in poorly done testing, look back at the old HardOCP reviews where that guy was such an Intel shill that Intel eventually offered him a Director of Marketing position and he shut down the site to take the job.

      I used to love HardOCP and his reviews but Jesus Christ could that guy slant a review in Intel's favor.
      Not really, a review should neutral report technical facts and observed performance.

      Comment


      • #13
        Originally posted by rene View Post

        Not really, a review should neutral report technical facts and observed performance.
        Depending on what you are reviewing, it may not be possible to be 100% objective.

        For example, if you review a game you can report on framerate objectively (though where in the game do you measure?). But what about bad gameplay? You could and should where possible motivate and describe such a statement of course (the gun handling feels "loose", the platforming is too floaty, the variety of items is too limited making the game monotonous), but there is no objective scale to measure on, and different people might have different opinions on such things.

        That comes in less in hardware reviews of course. But still, what about the typing feel on a laptop keyboard? That is near impossible to quantify, and even if you did (force profiles, travel distances, flex distances etc) that is not very useful because I don't know what a force of x millinewtons feels like, the number is meaningless to me as a reader.

        Similar applies to many other aspects, such as the overall impression of the product, taking into account objective and subjective analysis.

        Comment


        • #14
          Originally posted by rene View Post

          Not really, a review should neutral report technical facts and observed performance.
          Idealistic thinking.

          Reviews are written by human beings. Human beings are known to have biases. Those biases can sometimes appear in writing no matter how hard humans try to control themselves. So it is a never-ending problem.

          And others have said, and I agree, Michael runs an open forum, so I think you have go a long LONG way in your postings to offend him to the point of deleting your posts. IMHO your past posts do not show any signs of that extremism; I think you come across as one of the more level-headed people in these forums.

          Comment


          • #15
            Originally posted by rene View Post

            Not really, a review should neutral report technical facts and observed performance.
            The world is not black and white like this. It never has been. Even performance metrics can show unintentional biases in reporting due to various factors from misunderstanding what makes up a number, to bugs, to even misreporting irrelevant statistics. That doesn't even get into personal biases overt and covert. What I consider a relevant technical fact may be completely irrelevant to another writer or reader. I like factual reporting too, but simply stating a bunch of facts about a product is not a review nor is it journalism. That's just a spec sheet. A review is an editorializing of a person's experience with a product in the real world doing real work (or supposed to be anyway). It is by it's very nature an opinion and all opinions are subject to a person's POV biases. You can't have a proper review with out it. Otherwise you might as well just read the device's accompanied spec sheet for all the good it does.

            Comment


            • #16
              Originally posted by rene View Post

              Not really, a review should neutral report technical facts and observed performance.
              How do you review a distro solely with "technical facts"?

              If you just run benchmarks and give the version number of the included software, you haven't really done a review.

              In fact when Michael runs the Windows vs Linux vs BSD benchmarks, to me they are lacking, because they don;t tell you anything.

              OS 1 is able to encode video at x FPS and OS2 is able to encode in y FPS.

              Big Deal!!!

              It tell you nothing about the effort required to get the OS up and running and getting the encoding working.

              It tells you nothing of responsiveness of the OS during the encode.

              It tells you nothing about the ability play back the encode.

              Benchmarks are meaningless as far as I'm concerned.

              That s390x may be the fastest thing available and your work may be the best thing since sliced bread, but it does squat for me.

              Can I check my email, apply for a job, work on my resume, edit audio and video, play back video, play a game, work on a spreadsheet, work on my Python and FORTRAN programs, work on my custom distro, work on my websites, work on the AI program I am writing and more importantly how easy is it to get the system set up for this.

              Ubuntu based OSes are the easiest to set up from a clean install to so all these tasks.

              Run the custom script I have after a clean install, go get a soda, some back and everything is ready to go.

              That's what i care about.

              Comment


              • #17
                sophisticles

                Off topic: Is your username meant to rhyme with icicles, or Pericles?

                Comment


                • #18
                  Originally posted by Old Grouch View Post
                  sophisticles

                  Off topic: Is your username meant to rhyme with icicles, or Pericles?
                  with sophi's testicles

                  Comment


                  • #19
                    Originally posted by Old Grouch View Post
                    sophisticles

                    Off topic: Is your username meant to rhyme with icicles, or Pericles?
                    It is a homage to the Ancient Greeks, yes.

                    Comment


                    • #20
                      Originally posted by Gipr View Post

                      with sophi's testicles
                      Look who's back. You are the banned poster that has a thing about mayo on sandwiches, aren't you?

                      Michael, everyone's favorite banned poster is back.

                      Comment

                      Working...
                      X