Announcement

Collapse
No announcement yet.

NetBSD On The State & Future Of X.Org/X11

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

  • Originally posted by nia_netbsd View Post
    I've never seen such mastery in not reading the linked article before in my life, guys.

    Ultimately it's my choice to contribute to whatever open source projects I wish, and that happens to be netbsd and x.org rather than furthering the goals of The Linux Desktop. I work on BSD and X at my dayjob too for embedded systems.
    these people are just hateful, nothing else to say about it. they think they're entitled to your hard work, and if you aren't doing the right work for them you're a "problem". phoronix forums have no respect for individual autonomy, it's sad.

    Comment


    • Originally posted by avis View Post
      One 60Hz frame literally means 16.7ms which is the difference between a shitty mouse and a good mouse.
      Any decent compositor out there updates the mouse pointer independent of the rest of the screen. Whether you see tearing in windows or not has no effect whatsoever on your mouse.

      Comment


      • Originally posted by nia_netbsd View Post
        I've never seen such mastery in not reading the linked article before in my life, guys.

        Ultimately it's my choice to contribute to whatever open source projects I wish, and that happens to be netbsd and x.org rather than furthering the goals of The Linux Desktop. I work on BSD and X at my dayjob too for embedded systems.
        Part of the problem here is what is written in the article and re-quoted here..
        The bad news is that to have applications running we require access to a larger open source ecosystem, and that ecosystem has a lot of churn and is easily distracted by shiny new squirrels.
        Parties having different ideas where they are taking project in ways that do not suites your project should not be spoken about in this way. So this line is disrespectful to people right to contribute whatever they want.

        I am sorry to say NetBSD with X.org is going to have a hard road.

        Board members of x.org try finding a BSD one ever. I don't even know of a BSD person attempt to stand to be on the x.org board.

        Bare metal X11 at .x.org is in the hands of Alan Coopersmith. Fun he is not Linux developer he is a Solaris developer. This comes very clear when you have to raise particular security issues with him. The Unix desktop is still here that what BSD developers will be facing off against up-streaming.. This Unix desktop since it Solaris is highly conservative on changes that will be accepted. Yes life would have been better for up-streaming BSD work if Redhat was still in charge even better if it was a BSD person in charge..

        nia_netbsd you may not like it but you are fairly much looking what the Linux world was looking at in 1992-1993 when they wanted more graphical software resulting in a lot of bending of knees so it worked and introducing lots of things that were broken like sysvinit into Linux so they could have graphical stuff.

        Something to remember when Redhat was stopping doing bare metal they offer up the maintainer seat to anyone who could take it of x.org bare metal and no BSD person came forwards with the resources to take over the maintainer seat.

        nia_netbsd there is also the problem you have lot of people taking the idea since netbsd is going to keep X11 going everything alive on yourside that everything is good to be good.

        Yes that badly written line following is the way I would write it.

        The bad news is that to have applications running we require access to a larger open source ecosystem, and that ecosystem has a lot of churn and areas of ecosystem focusing on first wayland and areas of the ecosystem using new features that don't have is going to be harder to keep functional so requiring more developer resources.
        Yes this line does not read very upbeat the future of X11 in reality is not very upbeat in general without major changes.
        Last edited by oiaohm; 06 May 2024, 07:45 AM.

        Comment


        • Originally posted by avis View Post

          Wayland by default has a much higher input latency than Xorg due to "perfect frames", i.e. enforced vsync. That's especially bad for people with 60Hz monitors, let alone those who connect their 4K TV sets via HDMI 1.4b where only a 30Hz refresh rate is available.

          Another Wayland "advantage" busted hard.
          Much higher? Based on what? I'm playing games right now on a 144 fps monitor and I don't notice any significant latency. I do notice the smooth, perfect frames though.

          Comment


          • Originally posted by nia_netbsd View Post
            I've never seen such mastery in not reading the linked article before in my life, guys.

            Ultimately it's my choice to contribute to whatever open source projects I wish, and that happens to be netbsd and x.org rather than furthering the goals of The Linux Desktop. I work on BSD and X at my dayjob too for embedded systems.
            Then perhaps you should not have referred to other people who exercise their choice to contribute to whatever open source projects they wish (which happens to be furthering the goals of the Linux desktop rather than the BSD desktop) as “chasing shiny new squirrels”.

            Respect works both ways, or not at all.

            Comment


            • Originally posted by varikonniemi View Post

              Maybe your rage is blinding you so much that you did not study your enemy enough to know that if desired you can turn on tearing in wayland if you really need that worst case 17ms for placebo effect on a slow monitor.
              Reading comprehension for Wayland proponents is not just bad, it's missing altogether. Read my first message in this discussion and also enlighten us which Wayland compositors now support the tearing protocol. And before you say Mutter or Kwin supports it, read my first message again, again, and again. I don't bloody care that your $FAVOURITE_WAYLAND_COMPOSITOR supports it. The vast majority do NOT and have NO plans to even support it any time soon.

              Your protocol is A$$. Its implementations are an even worse A$$.

              And don't get me started on the fact that XFCE, the third most popular DE, 15 years after the protocol inception doesn't currently support Wayland at all.

              Blind zealotry and utter lack of knowledge are the two most outstanding traits of the Wayland proponents.

              Originally posted by krzyzowiec View Post

              Much higher? Based on what? I'm playing games right now on a 144 fps monitor and I don't notice any significant latency. I do notice the smooth, perfect frames though.
              CAN YOU READ SIR?

              In my message I was talking about 30fps TV sets and you're "refuting" my argument with your ... 144fps monitor. That's not a counter argument. I don't bloody care about your setup. Tons of people have their 4K TVs connected via HDMI 1.4b where only a 30Hz refresh rate is available. That's 33.33ms of extra latency for Christ's sake. Such monitors barely work under Xorg with no extra latency induced, under Wayland they are unusable.

              God, that's outright embarrassing. At least a dozen of people in this topic simply cannot read.

              "I feel so righteous about Wayland I don't care about your arguments!!! It works for me in KDE/Gnome!!!! I don't care that you have different HW and different Wayland compositor!!!! It works for me for on my 500Hz monitor I paid $2K for!!!!!!!! Oh and a brand new GPU worth at least $500!!!!Everyone must buy such monitors and GPUs because Wayland is cool!!!!!! I don't care that people maybe want to use their already existing HW! Or maybe they don't have the money. Wayland needs stuff to work!!!!"

              LMAO.
              Last edited by avis; 06 May 2024, 10:37 AM.

              Comment


              • Originally posted by avis View Post
                Tons of people have their 4K TVs connected via HDMI 1.4b where only a 30Hz refresh rate is available. That's 33.33ms of extra latency for Christ's sake. Such monitors barely work under Xorg with no extra latency induced, under Wayland they are unusable.
                What is the problem with a 30Hz monitor? It has 30Hz in X11 just as it has in wayland. The mouse does not care too much either way:-)

                Comment


                • avis: Chill dude

                  Comment


                  • Originally posted by tobias View Post

                    What is the problem with a 30Hz monitor? It has 30Hz in X11 just as it has in wayland. The mouse does not care too much either way:-)
                    Under Wayland you will care because you continue to prove you don't read, only write.

                    Wayland by default induces an extra frame latency for mouse movements which for a 30Hz refresh rate means your mouse movements will be randomly delayed by 0-33ms. Good luck with that.

                    Originally posted by access View Post
                    avis: Chill dude
                    I'm chill. It's +4C outside and roughly 18C in the room where I am. That was sarcasm, I know English well enough, I just went along with it.

                    Comment


                    • Originally posted by avis View Post
                      Wayland by default induces an extra frame latency for mouse movements which for a 30Hz refresh rate means your mouse movements will be randomly delayed by 0-33ms. Good luck with that.
                      This is not the case across all of wayland compositors and is not a problem with the Wayland protocol.
                      https://docs.kernel.org/gpu/drm-internals.html.
                      cursor plane
                      The mouse cursor is draw on it own output planes independent to graphical output from applications. Those two items are merged by the GPU as it sent to the monitor.

                      Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

                      Yes mouse cursor if compositor is done right is generated from its own CPU thread.

                      Avis some problems simple don't exist any more. The mouse cursor problem was implementation issue not Wayland protocol.

                      One of the advantages of the new drag and drop of windows in wayland is this in fact uses the cursor planes so resulting in lot lower CPU/GPU usage dragging and dropping windows so better performance.

                      Cursor planes work by what ever is the most current at scan out gets displayed. Yes you are running at 30hz output the cursor plane generation could be running at 8000hz. Cursor plane generation basically runs at what ever hz your mouse is if it working correctly. Yes not the busted gnome implementation that was fixed a few years ago.

                      Comment

                      Working...
                      X