LXQt 2.1.0 released with support for 7 Wayland sessions: Labwc, KWin, Wayfire, Hyprland, Sway, River and Niri

[-]

AndydeCleyre@reddit

Does anyone know if labwc is the only wayland window manager that can shade windows (show only the titlebar, temporarily)?

And can anyone comment on the flexibility of lxqt's panel?

[-]

RemarkableCuriosity@reddit

kwin_wayland can shade windows. This functionality is not exposed to the user but can be leveraged by KWin scripts

[-]

AndydeCleyre@reddit

Do you have an example or reference you can share?

[-]

RemarkableCuriosity@reddit

https://develop.kde.org/docs/plasma/kwin/api/ You can see that shade is a read/write property

[-]

AndydeCleyre@reddit

Thanks, but as far as I understand it doesn't actually work on Wayland.

[-]

Whoever's idea it was to not have a modernised version of Xorg and instead opt for window managers having to reimplement protocols or create new ones for core features (screen recording) making them incompatible with other wm messed up big time.

What a shitshow.

[-]

i_h8_yellow_mustard@reddit

modernised version of Xorg

Xorg hasn't been maintained in a long time. What exactly is the pathway for a "modern xorg"?

[-]

Dalcoy_96@reddit

Sorry, I meant X11 the protocol. Just take that protocol, remove all the depricated/old stuff, get rid of input (use libinput instead), make accessible using modern graphics APIs and give us a standard library across linux that developpers can plug into. The idea that I as an app developper need to seperatly implement different backends (screen sharing, scheenshot) for different wayland window managers is nuts.

This is all possible.

[-]

the_abortionat0r@reddit

Sorry, I meant X11 the protocol. Just take that protocol, remove all the depricated/old stuff,

So then we are back to ditching xorg.

Look, the actual devs of x have made it clear its unservicable. Everyone complains about Wayland breaking things but thats the whole reason x is being dropped in the first place. In order to fix x it would literally break things that already rely on it.

Whats the point of breaking x to "save "x instead of replacing it?

I'm tired of seeing the most technically ignorant people chanting the equivalent of "Screw Wayland just magically fix x!" as if thats even an option.

It has already been explained to death by REAL developers, the ones that LITERALLY maintained x for years and they them selves have pointed out exactly why your very childish request is impossible.

If x was savable some one would have forked it by now/actually done something.

Sure, people are trying to commit code but its not doing anything to save x as its literally not fixable.

This is all possible.

No, its not. You sound like a console kid saying the PS4 could do path tracing.

Its dead jim.

[-]

lproven@reddit

Look, the actual devs of x have made it clear its unservicable.

Again: no. Remember perspective.

The devs of the free hobbyist version don't want to do it.

X.org != X. X.org is one implementation out of dozens and dozens.

There are multiple other implementations for Linux alone.

https://www.xig.com/

There used to be multiple ones.

https://linux-center.org/en/X11/x11-servers/index.html

[-]

lproven@reddit

Just take that protocol, remove all the depricated/old stuff

I have been arguing for this for years now on the Register.

https://www.theregister.com/2022/06/22/the_x_window_system_just/

(use libinput instead)

WTF? No!

Look, X is the original cross platform GUI tool. It runs on many dozens of operating systems. I don't mean distros: Linux is Linux and it is 1 OS. You can futz around with C libraries and userlands but Linux is always Linux.

X runs on more entire OSes than Linux has had major kernel releases.

See this list?

https://www.opengroup.org/openbrand/register/

Every single one of those companies is paying over $110,000 every year to keep those OSes on that list. How many Linux distros have funding of > USD100K per annum? Very few!

There are, conservatively, 5-10x that many OSes still in active support and development that aren't on that list.

Windows runs X. All Windows back to Windows 2.0.

Classic MacOS and [all versions of Mac OS X. Plan 9. 9front. HarveyOS. JehanneOS. RISC OS. OpenVMS. MS-DOS, FFS. All the BSDs, including Free/Net/Open/Dragonfly/Ghost/Nomad/Midnight. Illumos. Tribblix. OpenIndiana.

Wayland is, by and large, a Linux toy. There are semi-serious ports to 2 other OSes. The Unix world is 10x bigger than Linux alone and that is only counting the ones that are in active commercial support, with tens to hundreds of millions being spent on them.

Always remember, the Unix world is 20 years older than Linux and it's many times bigger in terms of the amount of money spent on different products. No single one gets more spend than Linux except probably Apple, but put them all together and they make the Linux world look like a hobby project.

Which, never forget, it was and mostly still is.

[-]

daemonpenguin@reddit

This is false. X.Org is still maintained, the latest version came out about a week ago.

[-]

equeim@reddit

Separation between the display server and window manager brings its own issues. Though I'm not an expert so I can't say which approach is technically superior.

I also suspect that certain players in the game hoped that their DE would finally become "the Linux GUI" and other ones would die off, and therefore there would not be any fragmentation. Others simply underestimated the complexity of a fully-featured display server and how many additional protocols would be needed.

[-]

heretic_342@reddit

I wonder which Wayland session from the Cinanmon, XFCE, and LXQt is the most robust since all are considered experimental.

[-]

daemonpenguin@reddit

Xfce hasn't been officially released yet, so I probably wouldn't try that.

I have used Cinnamon's "experimental" Wayland session and it was better than either Plasma or GNOME's equivalents at the time.

[-]

Famous_Object@reddit

Cinnamon was better than Gnome and KDE Plasma? Both have been promoting their Wayland support as "almost ready" for years and Cinnamon just released an "experimental" session and it's already better?

Could you provide more details?

I know that Cinnamon is a Gnome fork and they can just piggy back on Gnome technology (before GTK 4 and libadwaita at least) and just add fixes for things Gnome overlooked but even so it's surprising.

[-]

daemonpenguin@reddit

Yes, Cinnamon offered a better experience, that's what I wrote. Yes, it was already better than GNOME and Plasma, despite being newer.

It was faster, more stable, fewer visual glitches, none of the usual mouse pointer flakiness.

Cinnamon doesn't just piggy back on GNOME. While it came from GNOME, it's a whole separate desktop these days with its own tools and approach.

[-]

Famous_Object@reddit

Do we even need Gnome now if Cinnamon can do everything it does but better? LOL

[-]

Fox3High369@reddit

Gnome and kde wayland are fully functional right now. I have used wayland since almost when it was included as an option for gnome and kde and barely encountered any issues.

wayland Cinnamon is incomplete and some apps wont even start. Maybe he has an Nvidia card?.

[-]

ppp7032@reddit

when was this testing?

[-]

daemonpenguin@reddit

First week of August, 2024.

[-]

Fox3High369@reddit

Do you have Nvidia?. I have been using gnome wayland since 2023 and no issues. Same for kde.

[-]

daemonpenguin@reddit

No, none of my machines have NVIDIA components. I run Intel and AMD cards.

I find it amusing Wayland proponents always have NVIDIA as their go-to boogie man. most Wayland implementations have plenty of problems across all the major video cards.

[-]

Fox3High369@reddit

I have gnome and kde wayland old laptop and new pc, no issues. I started using wayland since it first was available for both gnome and kde in 2023.

[-]

daemonpenguin@reddit

Wayland was available for both desktops long before 2023. It was mostly enabled by default in 2023. I've been testing Wayland periodically since at least 2018.

[-]

modified_tiger@reddit

LXQT should be solid, it just doesn't have a native compositor, but the parts have apparently been Wayland-ready for a while.

I need to try Cinnamon's wayland session, since I have a Mint install somewhere but if they're shipping it in Mint it should be at least minimally viable.

[-]

FunEnvironmental8687@reddit

That's because it never had its own window manager; it used Openbox instead. If you use LabWC, you'll notice very little difference in terms of functionality.

[-]

natermer@reddit

As far as desktop environments go... LXQT is going to be the actual "lite" one.

XFCE and Cinammon are often touted as "lightweight" alternatives to Gnome, but in practice it doesn't really work out. This was always sorta-true because the default Gnome desktops shipped by many distributions were bloated compared to Vanilla Gnome and this is where I think people get that idea from. Comparing apples to apples with Vanilla Gnome they still saved some memory out of the box, but once you had them all setup and running applications the savings are not that meaningful.

Lxqt, along with its precusor Lxde... offer a traditional Windows 9x style desktop setup with actual minimalism that is significant.

So it isn't in the same league as Cinnamon or XFCE and is kinda unfair to compare the two. If you want a full fledged desktop then LxQT is going to be spartan compared to those. If you want a minimalist desktop that is not tiling then LxQT with Labwc is probably the one you want to look at first.

And because of that minimalism, the benefit it gets from using QT libraries and KDE's wayland work, and ability to run on a variety simple Wayland compositors then I expect it is going to be the most mature one.

[-]

KnowZeroX@reddit

Isn't XFCE much lighter than Cinnamon and Gnome? Cinnamon was never really touted as lightweight, hence why Mint who makes Cinnamon offers MATE and XFCE options for those who want something more light weight.

Not disagreeing that LXQT is lighter than XFCE, that I agree with.

[-]

FunEnvironmental8687@reddit

There’s not much difference between modern XFCE and modern KDE in terms of performance. However, LXQt paired with LabWC or Sway will be more lightweight overall.

[-]

natermer@reddit

The thing to do is test yourself. Because a lot of it depends on what actual software you use.

When comparing Gnome to things like XFCE it is important to understand that Gnome ships with a lot of functionality and software that is designed to integrate together. Where as more typical "lite" environments you have to piece together.

This means that a lot of shared functionality is already loaded into memory for Gnome at bootup were as this doesn't happen as much for "lite" environments because they pull in software from all sorts of different projects with different libraries and dependencies and such.

So it is less important what the memory usage/etc is at start-up and more important what is memory usage/etc when you are actually using it. Like after you loaded up your browser, editor, terminal, or whatever set of software you actually use.

And, of course, it is all very dependent on what software you actually use. Like if you just load up Gnome desktop and don't use any of the Gnome software or features then that is going to be a pretty big waste no matter what.

Also there is a lot of dumb things Gnome does that uses gobs of memory sometimes which have relatively easy fixes. Like disabling the gnome-software daemons that check for updates and helps search for software to install. It is a nice feature, but it does use a lot of RAM sometimes.

I haven't tested XFCE in a long time, but depending on the exact setup I saw that it could save typical desktop user about 200-400MB of RAM, more or less.

That isn't meaningless, but it isn't that big of a deal when Google Chrome or Firefox will use 2-4GB of ram no problem.

I haven't looked closely at LxQT yet, but if it is anything like LXDE this is a type of desktop environment that you can run on systems with 128MB of ram or less. Which normal Linux systems won't even boot up anymore by default.

This is pretty significant because it is the type of desktop you can set in front of a normal PC user and they have a pretty good chance at figuring out how to use it. Which isn't going to happen with things like Sway or Hyprland.

I am looking forward to trying it out in Wayland form and confirming my suspicions.

[-]

oln@reddit

Both XFCE and LXQt are lightweight enough to run on my pentium 3 machine with 512 mb ram at least, using gentoo. I've booted it up with only 256 on XFCE as well (didn't test that with LXQt). Granted with a different distro where more features are compiled in and that uses systemd they might include a bit more.

Of course when you are running that memory constrained you have to be kinda conservative with what services you use and what features you enable, the average desktop distro is may enable a bunch of stuff out of the box (file indexing, network device discovery, internet account integration, printing services etc etc) even when using a light-weight DE that won't matter on a modern machine with 16 gb of ram but can make an impact on an old system.

[-]

FunEnvironmental8687@reddit

lxqt with labwc

[-]

bubblegumpuma@reddit

XFCE has to be compiled from git right now if you want the Wayland session. They have reached the 'feature freeze' point, though, IIRC, so if anyone wants to go through the pain, it should be fairly close to the release version - apart from bugfixes, of course.

[-]