Announcement

Collapse
No announcement yet.

Xfce's Xfwm4 Sees Wayland Port With Wlroots

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

  • #21
    Originally posted by skeevy420 View Post
    They need to be reran with only one DE installed on the system; not everything installed and then swapped between reboots since there is obvious spillover from that testing method that nulls these results.
    Yes, if that's the case, those tests are completely invalid. It goes without saying to check processes running etc., and this will tell you need to get rid of all installed packages of previously installed DEs. One can achieve that by e.g. creating package lists with pacman and delete all packages not included (so start with the DE you normally use).
    You can also mess things up further by starting Plasma via GDM etc., so imho just start DEs directly from tty.

    Comment


    • #22
      Originally posted by skeevy420 View Post

      kvuj

      That test is running Baloo on the Plasma session. That's not fair. It should be either turned off or allowed to finish before being rebooted and tested again. Then I noticed this:

      The GNOME session is ALSO RUNNING BALOO. It's also running kalendarac and kgpg. Those are all KDE services it probably shouldn't be running. Both the Mate and XFCE sessions are running all the aforementioned KDE stuff that GNOME is running.

      Those tests are clearly flawed. And that's just the stuff I noticed at a quick glace as a KDE user. For all I know there are accidental GNOME/Mate/XFCE stuff running in other places as well.

      They need to be reran with only one DE installed on the system; not everything installed and then swapped between reboots since there is obvious spillover from that testing method that nulls these results.
      Agreed. Also, the total RAM used in the process monitor doesn't match the "results" even slightly. Perhaps there's some confusion about virtual and physical memory going on here?

      I admit I can never remember the details of how memory management works, so I might be completely wrong, but IIRC it has something to do with the amount of memory actually used vs what is available for use by the process in those columns?

      Comment


      • #23
        Originally posted by hamishmb View Post
        I admit I can never remember the details of how memory management works, so I might be completely wrong, but IIRC it has something to do with the amount of memory actually used vs what is available for use by the process in those columns?
        I think total physical memory used is what can be considered crucial. Though oddly on Linux this can be influenced by some metrics that I'd consider cache, so use numbers directly after log-in or flush everything that can be flushed.

        Comment


        • #24
          And they said Xfce was a dead end and would not ever be ported to Wayland...
          Last edited by Vistaus; 15 July 2022, 10:58 AM.

          Comment


          • #25
            Originally posted by Anux View Post
            Thats an easy one, by beeing lightweight, working hassle free and not getting in your way.
            Sure, but there were and are so many WM's that function the exact same while also being lightweight. Don't get me wrong: I can see some of Xfce's appeal, but there's nothing in Xfwm that stands out enough to be viable enough.

            Comment


            • #26
              Not to complain about a large effort done without any monetary compensation, but I wonder about the future of Xfce. I just switched off it towards KDE, particularly because of the longstanding bug that makes dual monitors return to Mirror mode every time you leave suspend. This bug has been ongoing for months at least.

              KDE has been noted around 2 years back to consume about as much resources as Xfce while offering a much larger roster of services and much higher frequency of commits and growth. I have to say that I prefered Xfce for customisability over KDE, but the amount of low quality features in Xfce (visually dated, functional but not standardised, and such) is immediately visible compared to KDE, which basically offers the same but better in every field and for the same resources now.

              I don't wonder whether Xfce has a future or not. But rather about the "market" of Xfce. A DE will ultimately rely on its weight, features, and customisability for almost every Linux user. Xfce can't claim to succeed on the two former, and as for the latter, frankly I downloaded the same theme for KDE that I did for Xfce, I will probably replace some of my better monitoring tools with KDE widgets, and I consume about as much RAM.

              Is there a world where Xfce can stand right now? I feel like between the old-style window managers/ultra minimal DEs and the feature-packed beasts like KDE and Gnome, there is little room. I don't know if it's about development team size, availability, roadmap, or anything, but right now I can say that I used Xfce on my former computers, on this computer, and did so since about 2015. I switched out for KDE, and feel like I've lost nothing, including in performance, and gained a generally stronger and more modern experience. Wayland with 0 effort, Freesync activated through the DE...less bugs and just the same RAM/CPU requirements. Perhaps in a 2010 machine I'd see a bigger difference, but right now I really wonder what will Xfce has that others don't.

              Comment


              • #27
                I've mentioned in the past, but I am looking forward to someone coming up with a real lightweight shell, built from scratch (meaning, not catering to the past), Wayland-native, etc. Something like Sway maybe in lightness, but maybe able to do tiling and floating with a little more depth in decorations. Maybe the components are all layers (e.g. toolbar, notifications, possible top bar, etc.) Nice and simple config syntax that is already known, maybe TOML. I suppose someone could say, "hey, go make it happen", but unfortunately don't have these kinds of coding chops. That said, I am certain someone else out there has a similar itch and eventually will happen. I can picture it in my mind, just have no means of creating.

                Comment


                • #28
                  Originally posted by aufkrawall View Post
                  I think total physical memory used is what can be considered crucial. Though oddly on Linux this can be influenced by some metrics that I'd consider cache, so use numbers directly after log-in or flush everything that can be flushed.
                  Free RAM is there to be used, to test if a DE is capable of performing with lets say 1 GB of RAM you have to test on a machine with 1GB of RAM and not use 8 GB and see how much is left. From experience I can tell that XFCE is totally usable with 1GB, Gnome and KDE not so much.
                  Flushing memory or looking at the startup memory consumption doesn't give any indication on real life usage of a DE.

                  Comment


                  • #29
                    Originally posted by Vistaus View Post
                    Sure, but there were and are so many WM's that function the exact same while also being lightweight. Don't get me wrong: I can see some of Xfce's appeal, but there's nothing in Xfwm that stands out enough to be viable enough.
                    You are totally right, nothing special about Xfwm and nothing really bad too. It just happens to be the standard for XFCE. Its also easily replaceable but no one seems to care that much.

                    Originally posted by Vistaus View Post
                    I don't wonder whether Xfce has a future or not. But rather about the "market" of Xfce. A DE will ultimately rely on its weight, features, and customisability for almost every Linux user. Xfce can't claim to succeed on the two former, and as for the latter, frankly I downloaded the same theme for KDE that I did for Xfce, I will probably replace some of my better monitoring tools with KDE widgets, and I consume about as much RAM.
                    The RAM part I would certainly call bullshit (see my post above) but if you have over 2 GB and don't use RAM heavy applications you are probably fine with any DE and won't notice much difference. Also a difference of maybe 300 MB in RAM usage won't matter much if you use applications that need 16 GB or more.
                    And at last Xfwm is doing fine without hardware acceleration, that's a huge plus if you lack hardware or driver support (VM for example).

                    Is there a world where Xfce can stand right now? I feel like between the old-style window managers/ultra minimal DEs and the feature-packed beasts like KDE and Gnome, there is little room. I don't know if it's about development team size, availability, roadmap, or anything, but right now I can say that I used Xfce on my former computers, on this computer, and did so since about 2015. I switched out for KDE, and feel like I've lost nothing, including in performance, and gained a generally stronger and more modern experience. Wayland with 0 effort, Freesync activated through the DE...less bugs and just the same RAM/CPU requirements. Perhaps in a 2010 machine I'd see a bigger difference, but right now I really wonder what will Xfce has that others don't.
                    Feature wise, KDE and Gnome will never be matched. For me its just configurability and stability.
                    If someone used Gnome or KDE 10 years ago and tries to use it today its a completely different DE and user experience. For XFCE one would think, oh a new theme is being used (this might not be by design and more a lack of resources but the outcome is the same and I like it. Look at the home page: "release news: XFCE 4.16 from 2020" )
                    And I love to have my open/close buttons on the left side and my panel on top to reduce mouse movement.

                    Comment


                    • #30
                      I'm glad to see xfce adopting Wayland, the last of the 3 "main desktop environments" (maybe they should change to "wfce" now )

                      But the best part is that they opted for wlroots (which aims to be a more generic compositor than Mutter/Kwin) and I hope other projects like Mate and Lxde can work together with them to improve wlroots into a better compositor, rather than spending their limited resources in a more private solution.

                      Comment

                      Working...
                      X