• Re: [gentoo-user] Lots of issues with wayland

    From Wols Lists@21:1/5 to Daniel Frey on Mon Aug 5 19:00:02 2024
    On 05/08/2024 17:30, Daniel Frey wrote:
    1. Logins don't work reliably. I use KDE/SDDM and when logging in it
    appears to start on a new VT and sometime it doesn't start. Or it will
    start the new VT and fail to switch to it, leaving a text console and a blinking cursor. This didn't happen when using X11.

    Last I investigated, sddm had a *hard* dependency on X11. So even if
    you're running a Wayland system (like I am) you need X installed so that
    sddm will work.

    Daft, I know, but that seems to be the way things are at the moment ...

    Cheers,Wol

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From J. Aho@21:1/5 to Daniel Frey on Tue Aug 6 08:10:01 2024
    On 05/08/2024 18.30, Daniel Frey wrote:
    Is it just me or is wayland nowhere near primetime?

    There are still issues with wayland, not sure how up to date this page is: https://gist.github.com/probonopd/9feb7c20257af5dd915e3a9f2d1f2277

    My SailfishOS based phone uses wayland and on that it's been working
    fine, on my desktop with a nVidia RTX I haven't managed to even login
    into a wayland session, but on my laptop with i915 it works fine most of
    the time and it leave me think that the developers behind wayland are
    mainly using intel graphics cards and then blame the graphics driver
    when things don't work on other hardware.

    I would say for general use wayland ain't up for it's task, if you have
    the right hardware and you mainly use the browser, then it works.


    --
    //Aho

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Michael@21:1/5 to [email protected] on Tue Aug 6 10:17:04 2024
    On Tuesday, 6 August 2024 09:29:15 BST [email protected] wrote:
    On 05/08/2024 23:56, Daniel Frey wrote:
    I'm glad I'm not the only one having problems. It does make me wonder though if the discrete video card (nvidia) is the cause of some/most of these problems.

    If I were to 'guess', I think it's the Nvidia drivers and less to do
    with Wayland implementation.

    I have an Nvidia RTX 20-series and Wayland sessions generally exhibit
    mostly the same issues as yours - flickering and/or
    freezing/unresponsive panels, random KDE crashes, and general
    instability. Sleep simply doesn't work at all properly with a wayland
    session and my PC freezes, requiring a reboot. Mind you, this is a
    modern system with an X670E chipset, but certainly not modern enough to
    still have launch issues.

    It's the same with both the proprietary Nvidia driver and the new 1st
    party open-source driver that comes with the 555-series drivers. I've
    not tried nouveau - it's on my list, but I sometimes need CUDA so not
    yet sure how this is going to wor
    k out without a driver swap.

    They don't happen in X11 though.

    Same here. X11 is mostly stable for nvidia except for suspend/resume
    which can often not suspend at all, requiring multiple attempts, or
    borking up the screen resolution.

    I say 'guess' above in quotes because Nvidia have notoriously bad
    drivers for Linux and Wayland support has been pretty non-existent _and_
    I've been using Wayland as a daily driver with AMD (current laptop) and
    Intel GPUs (previous laptop) for about 2 years without any major issues outside of minor known KDE bugs which have got hoovered over time with updates.

    On various laptops and desktops of up to 15+ years old Wayland works here for some years now. They all have integrated and/or discrete radeon/AMD graphics.

    All PCs boot/suspend/wake up reliably. Bar the quite rare crash of kwin,
    which takes down Plasma but not any of the open application windows on it, I have not experienced any problems for quite a few years now. From what I
    recall X11 used to have more crashes, than what I am experiencing on Wayland.

    Besides Plasma I've also tried Enlightenment DE and it also mostly worked with Wayland.

    There is also an old 2008 laptop with Intel CPU and intel graphics which will not work fully with Wayland. It launches a desktop session fine, but some application windows (e.g. Firefox) will not respond to mouse clicks. It may
    be a matter of missing gtk packages/configuration, but I have not looked into it further.


    Furthermore, on my workstation (with the Nvidia GPU) all
    problems, Wayland and the X11 sleep issue , essentially disappear as
    soon as I wire things up to the onboard built-in Ryzen GPU and switch to
    the "amdgpu" driver and Wayland is rock solid - no sleep/resume issues either.

    So... yes, in my experience Nvidia is just crap on Linux and, frankly,
    given where they
    seem to be headed wrt their business priorities, I
    don't see this getting any better any time soon. I think my next GPU
    upgrade, whenever that may be, will be an AMD.

    I'm curious to see how KDE 6 will behave once stabilised. It's supposed
    to have further Wayland improvements, but I'm not holding my breath as I don't think it's necessarily KDE's problem.

    - Victor


    -----BEGIN PGP SIGNATURE-----

    iQIzBAABCAAdFiEEXqhvaVh2ERicA8Ceseqq9sKVZxkFAmax6hAACgkQseqq9sKV ZxkB/RAA3l0+o38KZyjfPuwXDOUmBfSYnkaGvYgOkYzZABwxtTYd/tsah8IzMI4M BGgp/Ffx7k8PUw5X5Ma3/E+wodcoB5xMWnWmBLa0WdHv6UPMrtdB0HwoU9/sNi8x TszME3TPJeTH10wfUABFOVawnMpxgqlxVpm7Zl2yOAdQLCftgPr1JfNDtYvxxQdS quKJy6OwlohftRAlqm5sVCaUmnZFlPF4gvFTtm9rHBJQsQnxMCeL0HaopdpfP6Fu GtWp1bWaNrbYYosd30hJJn09WLnpIiFhy2mRg6ylpz0sOGjp0PPeDT8ada9amh9u Q35/r+p74IiyZns2CiRTXCgLVtwuoPnCS8EQXuKlPI7qUI4zB9R8Q0D+6SLoTuK6 GBhh7s1NMihbKKS2wQbierBtKIvBNm1cLbnQAV1zba92/aFKQwHBsJ+M7SoKofqy OtPemk64yrR1BCTZ5fOw8pOxoBnZp2/xnJ9yfqz+f/gkflVX64Tm+54BZSW/AF8B mdv4YTChawpikCO71tcRCWwE3pmaW13wOG0PegfmANnigqK7t81+juxF8DF0e31V F8vBKijrGwevyPdDDEnDGjIvmLX3G23P/ijUmm64ghsAbl9rbEWYzlqRqW/A51YQ v2lQJiEsICAGkqVNa5+smNShDtujX9EHa/3EauXJtV+/HVnekcQ=
    =EsNx
    -----END PGP SIGNATURE-----

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Michael@21:1/5 to All on Tue Aug 6 11:30:19 2024
    On Monday, 5 August 2024 17:30:56 BST Daniel Frey wrote:
    Is it just me or is wayland nowhere near primetime?

    I did a switchover to systemd/wayland some time ago and it seemed to
    solve some problems I had.

    The problem is it also came with so many more issues than it fixed:

    1. Logins don't work reliably. I use KDE/SDDM and when logging in it
    appears to start on a new VT and sometime it doesn't start. Or it will
    start the new VT and fail to switch to it, leaving a text console and a blinking cursor. This didn't happen when using X11.

    I've come across such symptoms on more recent SDDM versions. SDDM would not launch, or would not launch fully on boot. Nothing meaningful in the logs. Strangely, it would start fine if I were to restart display-manager, but never on boot. This would only happen only on a ryzen PC, but not on other older
    and slower PCs.

    After a lot of testing I discovered the faster ryzen would finish starting all the openrc services including display-manager, *before* haveged had a chance
    to generate enough entropy. Hence SDDM failed to start. I have no TPM chip
    on this MoBo, so I don't know if it would help accelerate haveged or not.

    The solution was to add in /etc/rc.conf:

    rc_display_manager_after="local"

    to allow enough time for entropy generation by haveged before SDDM tries to launch.


    2. Sometimes video hardware acceleration just doesn't work. Now I do
    have a discrete nVidia card with the proprietary driver, but I switched
    to nouveau and it didn't work either. Again, not an issue in X11.

    I have a nVidia RTX 3070 Ti.

    From what has already been posted this seems like a Nvidia driver issue. Graphics acceleration works fine on AMD (both integrated and discrete
    graphics) with amdgpu or radeon drivers, depending on the age of the graphics cards.


    3. I sleep/resume a lot, wayland seems to be quite buggy on resume. As
    in, the other three issues listed here are far more likely to occur
    after a sleep/resume, but they do happen even after rebooting and being
    on the PC most of the day.

    There was a glitch a few months ago affecting one PC with an AMD APU here and
    a dual-monitor setup. The PC would sleep/resume OK, but some time later it would crash kwin if firefox was used heavily and was being dragged between the two monitors. Evidently a graphics driver issue on this integrated graphics PC. Following some recent xorg/mesa updates and enabling hardware
    acceleration on firefox has fixed this. All other AMD graphics systems have always been able to sleep/resume reliably.

    Again on this PC only, changing from the desktop to a console with Ctrl+Alt+F1 and back again would arrive at a distorted/full of tearing desktop. This may not happen the first time, but always happens after a couple of attempts. The workaround is to switch first to the VT where the SDDM was launched from and then to the VT where the desktop is running.


    4. This is the big one, the panel in KDE hangs intermittently. The clock
    will freeze and the entire panel is unresponsive (but weirdly enough,
    KDE doesn't realize it's not responding.) I've tried killing and
    restarting the process and that doesn't always fix it either as it
    becomes disconnected with any open processes and it's not possible to resolve. Either have to logout or reboot.

    Sounds like a kwin crash, which brings Plasma down. It can happen, although quite rare here. Check your ~/.local/share/sddm/wayland-session.log and or syslog. I haven't found a fix, perhaps restarting kwin could/would do it? I prefer to shut down the individual application windows still running and then restart display manager to get a new, clean, Plasma session, without anything hanging in the background.


    I've switched back to X11 for now and have been using it for over a
    week. I managed to figure out how to set up multimonitor for now (both
    on the KDE desktop and in SDDM) and I haven't had any of these issues.
    For the record, I typically use the computer a few times a day during
    the week and sleep/resume, and longer on weekends.

    After much googling and experimenting, I can't seem to make these issues
    go away... which got me wondering if someone else has experienced any of these.

    I have done hardware tests (load testing, RAM tests, etc) - all are clear.

    -Dan

    I think your issues are caused by how well your graphics can work with
    Wayland. I have no experience with Nvidia to know if this is fixable through some configuration tweaks. Given the direction of travel is toward Wayland, I would think Nvidia will eventually adjust their code to make it work with Wayland, or nouveau may get there first.
    -----BEGIN PGP SIGNATURE-----

    iQIzBAABCAAdFiEEXqhvaVh2ERicA8Ceseqq9sKVZxkFAmax+zsACgkQseqq9sKV Zxkq1w//WPtbwjaI0vFDCFjw51u3FCp1AUK22D3P71ZMkyyzElsaknIJKEwug6vG RtX6hjUiQk7YTihNRd4X9DWrWJq5KLY5lqM0bf5d5lK/uAUjeU3iBy1BUwtMpz9G MX9u5i8O6lvK2B8uVpcNPlUVOIj8ZJMlvesN1s27z8anKPV4Jea408iNa0uxhY+r YGzYFcH51f154QyMa4SMojVl+tv+AcczY4lnypgOVfDjkQgNAuTJPW3E+4l/+kKb X3mkqy82s03wOdd7CDC4twGoUODhC3/8SCIzs3yFVMCKhsoY0IT7MxASzUP7+oza gi7KOwrDgNFu6WtUqgokxpIEdQqe5AGxEYfsuAtUBa/fMS61VeNSZSmWrK29I/Py DeamKY5TvupvjXt5/qlSlcfvJOEj+y+7IhuMBFCDFmZa3YzoFvmcbFzn1FIcvRXi OhARzOqT0O6dXfo6xf71XLRKO4c5b8A/qAWyZ2RezyaEj1w262s4B55kL/kpHWeq SJhlhQsicKcLCGj5LoBpMa6Iw2uBw/t667A25tvAzZu3ayF1tdWWs1GexT1k4fZc CnFcypL9XcTppDuSBsdkEScdrsxxeuiAC5WwZuOldVf1/HuTDRLC8bwr3cEP0Dzn hERRyB3WTEq5nD67nYrkUY2AqF5g8qzyEcCQCpA5tEtNKE6mmpI=
    =ZSjb
    -----END PGP SIGNATURE-----

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From =?utf-8?Q?Arsen_Arsenovi=C4=87?=@21:1/5 to J. Aho on Tue Aug 6 16:00:01 2024
    "J. Aho" <[email protected]> writes:

    On 05/08/2024 18.30, Daniel Frey wrote:
    Is it just me or is wayland nowhere near primetime?

    There are still issues with wayland, not sure how up to date this page is: https://gist.github.com/probonopd/9feb7c20257af5dd915e3a9f2d1f2277

    it's either out of date or fearmongering. I'd presume the former if not
    for the first emphasised paragraph of the post: "DO NOT USE A WAYLAND
    SESSION! Let Wayland not destroy everything and then have other people
    fix the damage it caused. Or force more Red Hat/Gnome components (glib, Portals, Pipewire) on everyone".

    but, each "point" it makes can individually be debunked also. I'm not
    very interested in doing that beyond oneliners though (which is
    apparently enough for the author of the original post).

    - Wayland is broken by design - X is and has been for decades.
    - A crash in the window manager takes down all running applications -
    a crash in X takes down all running applications. only the label of
    what crashed differs
    - You cannot do a lot of things that you can do in Xorg by design -
    irrefutable as it is not specific enough (but, it is true that a lot
    of hacks X allows aren't permitted in wayland)
    - There is not one /usr/bin/wayland display server application that is
    desktop environment agnostic and is used by everyone (unlike with
    Xorg) - correct, this is also true of X WMs, and though people
    pretend all X desktops worked the same, anyone who has used a Java
    program, for instance, knows this is false
    the following point is the same

    - Wayland breaks screen recording applications - this is false, either
    1) screen recording works via portals, or 2) screen recording works
    via xwaylandvideobridge

    - Wayland breaks screen sharing applications - same as above

    - Wayland breaks automation software - see libei, I also have remote
    input working via kde connect, which relies on the same mechanisms
    automation would rely on

    - Wayland breaks Gnome-Global-AppMenu (global menus for Gnome) -
    possibly, I don't know, I don't use GNOME or global menus

    - Wayland broke global menus with KDE platformplugin - just tried krita,
    seems to works fine

    - Wayland breaks global menus with non-KDE Qt platformplugins - tried
    qtdesigner, seems to work fine

    - Wayland breaks AppImages that don't ship a special Wayland Qt plugin -
    I'm not sure what was anticipated - shipping outdated libraries will
    invariably break. note that Qt also doesn't work on X if you delete
    the XCB plugin.

    - Wayland breaks Redshift - this is true, but redshift isn't that
    special - I have redshift functionality in KDE, I had it in Sway, I
    know for a fact it's there in GNOME, just the redshift program is
    broken

    - Wayland breaks global hotkeys - they never worked, and continue not
    to, the reasoning for not allowing the hack X allows is quite good,
    see
    https://drewdevault.com/2019/02/10/Wayland-misconceptions-debunked.html

    - Wayland does not work for Xfce? -
    https://wiki.xfce.org/releng/wayland_roadmap

    - Wayland is biased toward Linux and breaks BSD - I've seen and worked
    on hobby OSes that implement Wayland just fine. obviously, porting is
    required (and it was required for X, eg OpenBSD still has homebrew X),
    but AFAIK freebsd runs wayland, and I know at least of one wayland
    desktop supports freebsd. it is ironic to bring up bsd when one of
    the arguments is "everyone has to reimplement basic functionality"
    given BSD kernel diversity, though

    - Wayland complicates server-side window decorations - CSD seems logical
    to me, at least in large part, so I'm not sure this actually matters.
    however, I know that SSD support exists.

    - Wayland breaks windows rasing/activating themselves - good! this is
    popup prevention, and doesn't actually impede *any* valid activation
    (window alerts still work, fresh apps windows will still come into
    focus, so do popups, etc, just fine)

    - Wayland breaks RescueTime - yes, because it relies on clients spying
    on eachother

    - Wayland breaks window managers - I'm not sure what they mean here

    - Wayland requires JWM, TWM, XDM, IceWM,... to reimplement Xorg-like
    functionality - libraries exist. note that WMs aren't no work either.
    anyone who has written one will be able to tell you that there are a
    lot of workarounds to implement to get WMs to work fine.

    - Wayland breaks _NET_WM_STATE_SKIP_TASKBAR protocol - unsure, really,
    literally never had to do that. i'll trust the electron devs on it
    not existing though

    - Wayland breaks NoMachine NX - I'm not sure what this program is
    either, it seems to be remote desktop, and according to the linked
    post it works now? /me shrugs - wayvnc exists anyway (for wlroots,
    IIRC KDE implements RDP)

    - Wayland breaks xclip - I was able to edit my clipboard with it, so I
    don't know what they mean. xwayland certainly forwards the clipboard

    - Wayland breaks SUDO_ASKPASS - what?

    - Wayland breaks X11 atoms - cars break horse saddles

    - Wayland breaks games - i played elden ring earlier today. there's
    this myth of forced double buffering on wayland that I presume is
    related, I'm not sure where it stems from but it's quite untrue. even
    adaptive sync on my freesync display worked fine. I do not know the
    intricacies of display syncing but I am 100% certain that double
    buffering is not forced

    - Wayland breaks xdotool - correct, actually; and unsurprisingly,
    similarly to rescuetime, xdotool relied on protocol flaws to work,
    libei exists to replace it

    - Wayland breaks xkill - not really surprising, again, since it exploits
    flaws in X. ISTR hitting a button accidentally that gave me a 'kill
    window' in my plasma wayland session, though, so I'm sure similar
    functionality exists

    - Wayland breaks screensavers - ironically, those never worked on X, but
    do on wayland, because the compositor can actually redirect keys
    properly rather than trying to patchwork around X

    - Wayland breaks setting the window position - yes, intentionally,
    wayland is declarative wrt window positioning (you might notice that
    if you right click, the popup position is as you'd expect - this is
    because the client tells the compositor where /relative to a surface/
    it wants some other surface to appear, rather than a global position).

    - Wayland breaks color mangement - did this ever work on X? unsure,
    but I know HDR was being worked on recently. my display is a regular
    RGB 24-bit display so I do not know how well it works.

    - Wayland breaks DRM leasing - decently sure this was implemented in
    wlroots, kde and gnome some years back, but I don't have the hardware
    to test.

    > "not all" highlights one of the fundamental problems with Wayland -
    > there is fragmentation

    not all X servers implement <XYZ>, not all X WMs implement <ABC>
    (e.g. again dwm doesn't implement EWMH). this isn't new

    - Wayland breaks In-home Streaming - I'm not sure what this means, but
    they cite steam remote play (together), a feature that I used on kde
    wayland, so there's that

    - Wayland breaks NetWM - so do X WMs (dwm for instance)

    - Wayland breaks window icons - this is true actually, I found it
    surprising but not particularly important since you can still have
    functional window icons (and I see a window icon on the very window
    I'm typing this in) by providing a desktop file

    - Wayland breaks drag and drop - I'm really not sure what to say here -
    I've used that feature.

    - Wayland breaks ./windowmanager --replace - X lacks --replace, but kwin
    and probably others don't

    - Wayland breaks Xpra - probably true but I never used Xpra so I can't
    be sure

    what these kinds of posts omit is that X is fundamentally broken and
    misfit to modern systems. the code is also unmaintained and not the
    greatest. the creation of wayland wasn't accidental, nor was it some
    ploy to undermine the desktop, nor ...

    there's a new graphics stack because the old one didn't work, and I'm
    not sure where this kind of response to it comes from - it seems similar
    to the systemd debacle. see also
    https://www.youtube.com/watch?v=GWQh_DmDLKQ (by a former X developer,
    so, someone with qualifications)

    what these posts /also/ lack is enthusiastic contributors who'll
    actually pick up the mantle of maintaining and improving X.

    I have been using wayland for probably four or so years now, and I've
    only so far missed a standard for global shortcuts (note, a _standard
    for it_, not a hack-job allowing clients to listen to keystrokes
    globally and react to them, leading to weird conflicts and fragmented configuration; something integrated into desktops, that's not a job of individual clients to implement, so, really, I'm missing that in X also).

    My SailfishOS based phone uses wayland and on that it's been working fine, on my desktop with a nVidia RTX I haven't managed to even login into a wayland session,

    hmm, on my 1050Ti PC I run kde plasma just fine. I don't know anyone
    with an RTX card to ask them if it works for them, though - but it could
    be some misconfiguration (you need some USE flags on the nvidia drivers
    IIRC, and some compositors, like sway, require an extra flag).

    but on my laptop with i915 it works fine most of the time and it leave
    me think that the developers behind wayland are mainly using intel
    graphics cards and then blame the graphics driver when things don't
    work on other hardware.

    not really - all the drivers in mesa are supported by most developers,
    and frequently tested. I actively use NVidia, AMD and Intel cards, and
    out of those, NVidia is the only one with occasional bugs (which are, no
    doubt, in the driver..)

    I would say for general use wayland ain't up for it's task, if you have the right hardware and you mainly use the browser, then it works.

    i don't have the right hardware (proprietary nvidia card as i
    mentioned), nor do i primarily use the browser (i work a lot of gtk and
    qt programs, and I know wine also works fine, probably through
    xwayland), so I have to disagree.

    have a lovely day
    --
    Arsen Arsenović

    --=-=-Content-Type: application/pgp-signature; name="signature.asc"

    -----BEGIN PGP SIGNATURE-----

    iOcEARYKAI8WIQT+4rPRE/wAoxYtYGFSwpQwHqLEkwUCZrIrmF8UgAAAAAAuAChp c3N1ZXItZnByQG5vdGF0aW9ucy5vcGVucGdwLmZpZnRoaG9yc2VtYW4ubmV0RkVF MkIzRDExM0ZDMDBBMzE2MkQ2MDYxNTJDMjk0MzAxRUEyQzQ5MxEcYXJzZW5AZ2Vu dG9vLm9yZwAKCRBSwpQwHqLEk3+EAP0e+RWzVbCppLolW8lrVCCIG8FIaXalg8xl cyW6SH4ywQEAqbbcAFkr6zO4gMbV20KpT3Wh9n01EvO+QvEdt0oemAQ=vSG+
    -----END PGP SIGNATURE-----

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From =?utf-8?Q?Arsen_Arsenovi=C4=87?=@21:1/5 to Wols Lists on Tue Aug 6 16:10:01 2024
    Wols Lists <[email protected]> writes:

    On 05/08/2024 17:30, Daniel Frey wrote:
    1. Logins don't work reliably. I use KDE/SDDM and when logging in it appears >> to start on a new VT and sometime it doesn't start. Or it will start the new >> VT and fail to switch to it, leaving a text console and a blinking
    cursor. This didn't happen when using X11.

    Last I investigated, sddm had a *hard* dependency on X11. So even if you're running a Wayland system (like I am) you need X installed so that sddm will work.

    Daft, I know, but that seems to be the way things are at the moment ...

    there's a draft somewhere that changes that dep, but upstream support is experiemntal so it isn't in-tree yet. hopefully SDDM comes under the
    KDE umbrella and maintenance picks up, though.
    --
    Arsen Arsenović

    --=-=-Content-Type: application/pgp-signature; name="signature.asc"

    -----BEGIN PGP SIGNATURE-----

    iOcEARYKAI8WIQT+4rPRE/wAoxYtYGFSwpQwHqLEkwUCZrItH18UgAAAAAAuAChp c3N1ZXItZnByQG5vdGF0aW9ucy5vcGVucGdwLmZpZnRoaG9yc2VtYW4ubmV0RkVF MkIzRDExM0ZDMDBBMzE2MkQ2MDYxNTJDMjk0MzAxRUEyQzQ5MxEcYXJzZW5AZ2Vu dG9vLm9yZwAKCRBSwpQwHqLEk31BAQDjbdRy6NICeLWx0Y2nAW+UAgaHKOKp4vZ5 ervdin7SwgD/aSakeORwIM+wiWbdbytxCCsjlgBYqq7wQILtj05riAY=Vaow
    -----END PGP SIGNATURE-----

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From =?utf-8?Q?Arsen_Arsenovi=C4=87?=@21:1/5 to Daniel Frey on Tue Aug 6 16:10:01 2024
    Daniel Frey <[email protected]> writes:

    Is it just me or is wayland nowhere near primetime?

    I did a switchover to systemd/wayland some time ago and it seemed to solve some
    problems I had.

    The problem is it also came with so many more issues than it fixed:

    1. Logins don't work reliably. I use KDE/SDDM and when logging in it appears to
    start on a new VT and sometime it doesn't start. Or it will start the new VT and fail to switch to it, leaving a text console and a blinking cursor. This didn't happen when using X11.

    hmm, please check journalctl / the wayland session log. I've never seen
    that and it more likely than not indicates KWin or SDDM crashing.

    2. Sometimes video hardware acceleration just doesn't work. Now I do have a discrete nVidia card with the proprietary driver, but I switched to nouveau and
    it didn't work either. Again, not an issue in X11.

    I have a nVidia RTX 3070 Ti.

    I'd expect the proprietary driver to work better. what kind of hardware acceleration do you mean however? 3D acceleration always worked for me
    but I don't recall if I tried video acceleration

    3. I sleep/resume a lot, wayland seems to be quite buggy on resume. As in, the
    other three issues listed here are far more likely to occur after a sleep/resume, but they do happen even after rebooting and being on the PC most
    of the day.

    this is IIRC a bug in nvidia drivers, I've had it happen on my desktop
    PC (1050Ti), but never on my AMDGPU desktop or Intel laptop.

    4. This is the big one, the panel in KDE hangs intermittently. The clock will freeze and the entire panel is unresponsive (but weirdly enough, KDE doesn't realize it's not responding.) I've tried killing and restarting the process and
    that doesn't always fix it either as it becomes disconnected with any open processes and it's not possible to resolve. Either have to logout or reboot.

    'plasmashell --replace' always fixed this on my nvidia machine. I've
    never been able to reproduce it off of it. decently sure that's https://bugs.kde.org/show_bug.cgi?id=469016 though (check, please, and
    let the devs know if it isn't)

    I've switched back to X11 for now and have been using it for over a week. I managed to figure out how to set up multimonitor for now (both on the KDE desktop and in SDDM) and I haven't had any of these issues. For the record, I typically use the computer a few times a day during the week and sleep/resume,
    and longer on weekends.

    After much googling and experimenting, I can't seem to make these issues go away... which got me wondering if someone else has experienced any of these.

    I have done hardware tests (load testing, RAM tests, etc) - all are clear.

    it is unlikely to be a hw fault - nvidia drivers are quite bad,
    especially at wayland, since nvidia mostly cares about compute on linux.

    on nvidia hw, using X is often the better choice unfortunately.

    hope that helps, have a lovely day.
    --
    Arsen Arsenović

    --=-=-Content-Type: application/pgp-signature; name="signature.asc"

    -----BEGIN PGP SIGNATURE-----

    iOcEARYKAI8WIQT+4rPRE/wAoxYtYGFSwpQwHqLEkwUCZrIszl8UgAAAAAAuAChp c3N1ZXItZnByQG5vdGF0aW9ucy5vcGVucGdwLmZpZnRoaG9yc2VtYW4ubmV0RkVF MkIzRDExM0ZDMDBBMzE2MkQ2MDYxNTJDMjk0MzAxRUEyQzQ5MxEcYXJzZW5AZ2Vu dG9vLm9yZwAKCRBSwpQwHqLEk6hMAQDnYT0D1pPRUFgqvPRNZuCIRZcb9K5hhNhQ kJvqPL3KbAD/fyNKj2Yfv/CggSsxY1a9KcaZhYew05XveOwI+gu2oQI=dh9e
    -----END PGP SIGNATURE-----

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Michael@21:1/5 to All on Tue Aug 6 14:57:09 2024
    On Monday, 5 August 2024 23:56:02 BST Daniel Frey wrote:
    On 8/5/24 09:58, Wols Lists wrote:
    On 05/08/2024 17:30, Daniel Frey wrote:
    1. Logins don't work reliably. I use KDE/SDDM and when logging in it
    appears to start on a new VT and sometime it doesn't start. Or it will
    start the new VT and fail to switch to it, leaving a text console and
    a blinking cursor. This didn't happen when using X11.

    Last I investigated, sddm had a *hard* dependency on X11. So even if
    you're running a Wayland system (like I am) you need X installed so that sddm will work.

    Daft, I know, but that seems to be the way things are at the moment ...

    Cheers,Wol

    I even tried enabling wayland for sddm (which is apparently
    experimental) but it didn't help either.

    Both X11 and wayland appear to be installed.

    I have not been able to make SDDM to launch rootless, either with x11-user, or wayland (configured to use kwin). On both of these two settings SDDM tries to launch on VT1 instead of VT7, which is occupied by the boot process agetty and consequently blocks SDDM from starting:

    (EE) HELPER: Failed to take control of "/dev/tty1" ("root"): Operation not permitted

    I see there's a fix for this:

    https://github.com/sddm/sddm/pull/1896

    but I don't think it has arrived in portage yet, or if it has it still won't work here on stable Plasma on openrc and no plymouth.

    I just find it odd that a lot of DEs (and distros) are defaulting to
    wayland when it's horribly broken.

    Dan


    -----BEGIN PGP SIGNATURE-----

    iQIzBAABCAAdFiEEXqhvaVh2ERicA8Ceseqq9sKVZxkFAmayK7UACgkQseqq9sKV ZxmvbhAA4GI8v9fEu1Glosn3f4aYNbmNxGE/CKEa3N+CTSVwIInyt2vFOopvLCD4 DrB9+4wF+9+KgsaoADvuJtCUl9+BS6m8RvtBnkz53aYpYTINXqX+cug/++xcHb8y a22wVWuL6bC8cRyTXrJG0IcAa5oGvUb77khXRl1nNtpx16aozzk9N0lG8Smj6joy LgY8b1esJtS2IUl3od2WmeMmGczic6a0CZUIq2XJDDxsrAqRwFE9Ddv6unGoFqQZ eXNBTWu8o/iwK/YZa37J4w31OYO16avyYe5orundGW2NILOvh8FhkZg8Ts+yR2O0 hpZxXP6NziwX60vJmQCvbi/E44gIeTbbpYLuPwYPLQsyM4/uYTeuHem4yUoXbIEI +1AyH6F0kGtGp53Yq0Q9NNHj2nsv/DGdepSAxEudGMULX28VupMgQfFxsf3747Ys 8r9c1nCt4s2UBBHMpozbI99ir0/xrDw8JtrTyEV4qfIGw9bT47SqVV0+DB2eBY+h bImetDhCcQsjuyuPezjpxKVI3RUe2xLG2zzOcKDjAS5KL4cPAh3O7s92d+SNPxP9 u/iH+a/y1f2XVI9mPKUV129LRKKJRq3hlvmjwknMiK1Gg+ENrZ80la6AiZp81fOz noBxJ15mv4ZR4dj2z22hg56s9tGKSTj6EDiK4ccD00DfhGKebOw=
    =6CLI
    -----END PGP SIGNATURE-----

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Alan Mackenzie@21:1/5 to All on Tue Aug 6 20:40:01 2024
    Hello, Arsen.

    On Tue, Aug 06, 2024 at 15:56:40 +0200, Arsen Arsenović wrote:
    "J. Aho" <[email protected]> writes:

    On 05/08/2024 18.30, Daniel Frey wrote:
    Is it just me or is wayland nowhere near primetime?

    There are still issues with wayland, not sure how up to date this page is: https://gist.github.com/probonopd/9feb7c20257af5dd915e3a9f2d1f2277

    it's either out of date or fearmongering. I'd presume the former if not
    for the first emphasised paragraph of the post: "DO NOT USE A WAYLAND SESSION! Let Wayland not destroy everything and then have other people
    fix the damage it caused. Or force more Red Hat/Gnome components (glib, Portals, Pipewire) on everyone".

    but, each "point" it makes can individually be debunked also. I'm not
    very interested in doing that beyond oneliners though (which is
    apparently enough for the author of the original post).

    I'm interested in one of the points you address in particular, namely:

    [ .... ]

    - Wayland breaks setting the window position - yes, intentionally,
    wayland is declarative wrt window positioning (you might notice that
    if you right click, the popup position is as you'd expect - this is
    because the client tells the compositor where /relative to a surface/
    it wants some other surface to appear, rather than a global position).

    This has caused problems for Emacs, though I can't remember exactly how,
    so I'm guessing. On restarting a saved Emacs session, it's necessary to
    have the windows the same size, in the same place they were on saving
    the session. This would appear to be difficult in Wayland.

    You say "yes, intentionally, wayland is declarative wrt window
    positioning". What does that mean when you replace abstract words like "declarative" with concrete sentences? What is declaring what to what
    else in this context, and what does that have to do with not being able
    to position windows?

    Later on, you say "where /relative to a surface/". I think "surface" is
    a word with particular meaning in Wayland, and using it in a Gentoo list without explanation is less than helpful. What does "surface" mean in
    this context? Is the entire screen such a "surface"?

    So, is it possible in Wayland to record a configuration of windows,
    their sizes and positions, then restore these on starting a program
    again? If not, that would appear to be a design bug in Wayland. What
    am I missing?

    [ .... ]

    have a lovely day
    --
    Arsen Arsenović

    --
    Alan Mackenzie (Nuremberg, Germany).

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From J. Aho@21:1/5 to All on Tue Aug 6 22:40:01 2024
    On 06/08/2024 15.56, Arsen Arsenović wrote:
    "J. Aho" <[email protected]> writes:

    On 05/08/2024 18.30, Daniel Frey wrote:
    Is it just me or is wayland nowhere near primetime?

    There are still issues with wayland, not sure how up to date this page is: >> https://gist.github.com/probonopd/9feb7c20257af5dd915e3a9f2d1f2277

    it's either out of date or fearmongering. I'd presume the former if not
    for the first emphasised paragraph of the post: "DO NOT USE A WAYLAND
    SESSION

    It's quite possible as I didn't really read the beginning of the page,
    just saw some stuff mentioned that I had seen on other places.

    - Wayland breaks screen recording applications - this is false, either
    1) screen recording works via portals, or 2) screen recording works
    via xwaylandvideobridge

    I think there was some merit to that claim some years ago, I'm not into
    screen recording so not sure.


    - Wayland complicates server-side window decorations - CSD seems logical
    to me, at least in large part, so I'm not sure this actually matters.
    however, I know that SSD support exists.

    I have to say I'm quite anti-CSD myself, much for the lack of having
    close on left side and minimize/maximize on the right side of the window
    but also the horror of everyone want to make their own look on the CSD
    which makes everything to look like microsoft-windows where even
    microsoft's own application uses different look.

    I don't see the CSD as a specific wayland issue, it has the same issue
    in X11 too.



    - Wayland breaks window managers - I'm not sure what they mean here

    I guess that could be related to X11 window managers that do not support wayland, but it's like a qt3 application don't run on a system with only
    qt6 libraries.


    there's a new graphics stack because the old one didn't work, and I'm
    not sure where this kind of response to it comes from - it seems similar
    to the systemd debacle. see also
    https://www.youtube.com/watch?v=GWQh_DmDLKQ (by a former X developer,
    so, someone with qualifications)

    what these posts /also/ lack is enthusiastic contributors who'll
    actually pick up the mantle of maintaining and improving X.

    Sure it's far easier to say what "you" think don't work than code.
    There was a X11 variant where they had stripped away loads of support
    for old stuff, just don't remember the name (most variants of X11 had
    been forgotten), and they had some improvements too, but if you don't
    get picked up by a major distribution, you will definitely have a lot
    harder to break thru to the masses.


    I have been using wayland for probably four or so years now, and I've
    only so far missed a standard for global shortcuts (note, a _standard
    for it_, not a hack-job allowing clients to listen to keystrokes
    globally and react to them, leading to weird conflicts and fragmented configuration; something integrated into desktops, that's not a job of individual clients to implement, so, really, I'm missing that in X also).

    Just make a patch and submit it ;)


    My SailfishOS based phone uses wayland and on that it's been working fine, on
    my desktop with a nVidia RTX I haven't managed to even login into a wayland >> session,

    hmm, on my 1050Ti PC I run kde plasma just fine. I don't know anyone
    with an RTX card to ask them if it works for them, though - but it could
    be some misconfiguration (you need some USE flags on the nvidia drivers
    IIRC, and some compositors, like sway, require an extra flag).

    I have been using KDE since Havoc called me a troll for pointing out
    that you can't have configuration options that has the exact same name
    but does different things.
    I have rebuilt the whole system, I have tried with other distros, the
    same result, I do not get wayland to work on the nVidia, not sure how
    many different howto's I have tried to follow, but on the laptop not
    much of an issue at all but as I pointed out it's an intel GPU. And my
    old mobile it works fine too.

    --
    //Aho

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From =?utf-8?Q?Arsen_Arsenovi=C4=87?=@21:1/5 to J. Aho on Tue Aug 6 23:50:01 2024
    "J. Aho" <[email protected]> writes:

    On 06/08/2024 15.56, Arsen Arsenović wrote:
    "J. Aho" <[email protected]> writes:

    On 05/08/2024 18.30, Daniel Frey wrote:
    Is it just me or is wayland nowhere near primetime?

    There are still issues with wayland, not sure how up to date this page is: >>> https://gist.github.com/probonopd/9feb7c20257af5dd915e3a9f2d1f2277
    it's either out of date or fearmongering. I'd presume the former if not
    for the first emphasised paragraph of the post: "DO NOT USE A WAYLAND
    SESSION

    It's quite possible as I didn't really read the beginning of the page, just saw
    some stuff mentioned that I had seen on other places.

    - Wayland breaks screen recording applications - this is false, either
    1) screen recording works via portals, or 2) screen recording works
    via xwaylandvideobridge

    I think there was some merit to that claim some years ago, I'm not into screen
    recording so not sure.

    While xwaylandvideobridge is a relatively new innovation, portals
    aren't, and the post would appear to be receiving updates, so the
    omission is strange.

    - Wayland complicates server-side window decorations - CSD seems logical
    to me, at least in large part, so I'm not sure this actually matters.
    however, I know that SSD support exists.

    I have to say I'm quite anti-CSD myself, much for the lack of having close on left side and minimize/maximize on the right side of the window but also the horror of everyone want to make their own look on the CSD which makes everything to look like microsoft-windows where even microsoft's own application uses different look.

    I don't see the CSD as a specific wayland issue, it has the same issue in X11 too.

    Yes, this is also one of my reasons for preferring SSD (which KDE does I believe).

    - Wayland breaks window managers - I'm not sure what they mean here

    I guess that could be related to X11 window managers that do not support wayland, but it's like a qt3 application don't run on a system with only qt6 libraries.

    Possibly.


    there's a new graphics stack because the old one didn't work, and I'm
    not sure where this kind of response to it comes from - it seems similar
    to the systemd debacle. see also
    https://www.youtube.com/watch?v=GWQh_DmDLKQ (by a former X developer,
    so, someone with qualifications)
    what these posts /also/ lack is enthusiastic contributors who'll
    actually pick up the mantle of maintaining and improving X.

    Sure it's far easier to say what "you" think don't work than code.
    There was a X11 variant where they had stripped away loads of support for old stuff, just don't remember the name (most variants of X11 had been forgotten),
    and they had some improvements too, but if you don't get picked up by a major distribution, you will definitely have a lot harder to break thru to the masses.

    That's indeed the case, XDG and freedesktop and the X consortium being
    behind Wayland certainly helped adoption.

    I have been using wayland for probably four or so years now, and I've
    only so far missed a standard for global shortcuts (note, a _standard
    for it_, not a hack-job allowing clients to listen to keystrokes
    globally and react to them, leading to weird conflicts and fragmented
    configuration; something integrated into desktops, that's not a job of
    individual clients to implement, so, really, I'm missing that in X also).

    Just make a patch and submit it ;)

    I was actually in the process of doing that when I discovered a portal
    one did exist already; https://flatpak.github.io/xdg-desktop-portal/docs/doc-org.freedesktop.portal.GlobalShortcuts.html

    IDR what the status of adoption was at the time, but I left it at that
    for developers to pick up at their own pace.

    My SailfishOS based phone uses wayland and on that it's been working fine, on
    my desktop with a nVidia RTX I haven't managed to even login into a wayland >>> session,
    hmm, on my 1050Ti PC I run kde plasma just fine. I don't know anyone
    with an RTX card to ask them if it works for them, though - but it could
    be some misconfiguration (you need some USE flags on the nvidia drivers
    IIRC, and some compositors, like sway, require an extra flag).

    I have been using KDE since Havoc called me a troll for pointing out that you can't have configuration options that has the exact same name but does different things.
    I have rebuilt the whole system, I have tried with other distros, the same result, I do not get wayland to work on the nVidia, not sure how many different
    howto's I have tried to follow, but on the laptop not much of an issue at all but as I pointed out it's an intel GPU. And my old mobile it works fine too.

    That's quite unfortunate - I don't have lots of Nvidia hardware or
    expertise to help with that, my apologies. I've only tested on the
    dinky 1050Ti I have, and I've been using it for Wayland since Nvidia
    added GBM support to their beta driver (it was /very/ awful back then),
    so, not a lot of variety.
    --
    Arsen Arsenović

    --=-=-Content-Type: application/pgp-signature; name="signature.asc"

    -----BEGIN PGP SIGNATURE-----

    iIcEARYKAC8WIQT+4rPRE/wAoxYtYGFSwpQwHqLEkwUCZrKYNREcYXJzZW5AZ2Vu dG9vLm9yZwAKCRBSwpQwHqLEkwKaAP0c3xVn84cn+wkI3hb4wevsyZWOxdUAnNde QyebnLEQogEAqccTVl/4FiyLJKVkKB9qhB2TWdl9E4oer2gNJ16bFQ8�0z
    -----END PGP SIGNATURE-----

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wol@21:1/5 to Alan Mackenzie on Wed Aug 7 00:10:02 2024
    On 06/08/2024 19:31, Alan Mackenzie wrote:
    So, is it possible in Wayland to record a configuration of windows,
    their sizes and positions, then restore these on starting a program
    again? If not, that would appear to be a design bug in Wayland. What
    am I missing?

    That - unlike X - is because windows cannot say where they are going to
    go. They can *ask* where they want to go, which isn't the same thing.

    Iirc, X behaves like Windows, which means applications can *seize*
    focus, which drives me up the wall on occasion at work. I'll have an
    Excel macro running, which takes maybe 3 or 4 minutes. So I go into
    let's say Slack. Excel triggers something (google drive?) which grabs
    focus and disappears, so all of a sudden I *think* I'm gaily typing into
    Slack. But focus has been stolen and I'm typing into a vacuum -
    EXTREMELY frustrating, especially as I don't actually know what's going on.

    In Wayland, you can't steal focus. But as a side effect, it's Wayland
    that controls the window, not the application. So Wayland is more
    secure, but that comes with unavoidable side effects that you don't like...

    Cheers,
    Wol

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From =?utf-8?Q?Arsen_Arsenovi=C4=87?=@21:1/5 to Daniel Frey on Tue Aug 6 23:30:01 2024
    Daniel Frey <[email protected]> writes:

    On 8/6/24 06:56, Arsen Arsenović wrote:
    "J. Aho" <[email protected]> writes:
    - Wayland breaks screensavers - ironically, those never worked on X, but
    do on wayland, because the compositor can actually redirect keys
    properly rather than trying to patchwork around X

    I actually forgot about this one - the screensaver was crashing all the time causing me to use loginctl to unlock the session.

    That hasn't happened even once in the last 8 days since using X11.

    Possibly https://bugs.kde.org/show_bug.cgi?id=489072 or https://bugs.kde.org/show_bug.cgi?id=489180 (so, a nasty coincidence).

    Dan

    --
    Arsen Arsenović

    --=-=-Content-Type: application/pgp-signature; name="signature.asc"

    -----BEGIN PGP SIGNATURE-----

    iIcEARYKAC8WIQT+4rPRE/wAoxYtYGFSwpQwHqLEkwUCZrKVjxEcYXJzZW5AZ2Vu dG9vLm9yZwAKCRBSwpQwHqLEk7LJAP4sxeK06IcW3khs60Vlll/zi/jdqSd0Z92Z ooLjedGgnwD+LjNFk1kM9FDeUbn6vtGb3dViZbv7ZishPFJGTzA5iQY�tT
    -----END PGP SIGNATURE-----

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wol@21:1/5 to All on Wed Aug 7 00:20:01 2024
    On 06/08/2024 22:40, Arsen Arsenović wrote:
    That's indeed the case, XDG and freedesktop and the X consortium being
    behind Wayland certainly helped adoption.

    Well, Wayland is - effectively - X13.

    Cheers,
    Wol

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From =?utf-8?Q?Arsen_Arsenovi=C4=87?=@21:1/5 to Alan Mackenzie on Wed Aug 7 00:20:01 2024
    hi Alan,

    thank you for maintaining CC mode :-) it is endlessly useful in my
    day-to-day.

    to preface, I'm not a wayland protocol expert; I've had to debug clients
    and servers a few times, view communications between clients and servers
    some others, experimented with some patches, but I haven't written a significant amount of code with it, and am mostly an end user happy to
    see an end to flickering, window tearing, weird keyboard grabs, ...

    Alan Mackenzie <[email protected]> writes:

    Hello, Arsen.

    On Tue, Aug 06, 2024 at 15:56:40 +0200, Arsen Arsenović wrote:
    "J. Aho" <[email protected]> writes:

    On 05/08/2024 18.30, Daniel Frey wrote:
    Is it just me or is wayland nowhere near primetime?

    There are still issues with wayland, not sure how up to date this page is: >> > https://gist.github.com/probonopd/9feb7c20257af5dd915e3a9f2d1f2277

    it's either out of date or fearmongering. I'd presume the former if not
    for the first emphasised paragraph of the post: "DO NOT USE A WAYLAND
    SESSION! Let Wayland not destroy everything and then have other people
    fix the damage it caused. Or force more Red Hat/Gnome components (glib,
    Portals, Pipewire) on everyone".

    but, each "point" it makes can individually be debunked also. I'm not
    very interested in doing that beyond oneliners though (which is
    apparently enough for the author of the original post).

    I'm interested in one of the points you address in particular, namely:

    [ .... ]

    - Wayland breaks setting the window position - yes, intentionally,
    wayland is declarative wrt window positioning (you might notice that
    if you right click, the popup position is as you'd expect - this is
    because the client tells the compositor where /relative to a surface/
    it wants some other surface to appear, rather than a global position).

    This has caused problems for Emacs, though I can't remember exactly how,
    so I'm guessing. On restarting a saved Emacs session, it's necessary to
    have the windows the same size, in the same place they were on saving
    the session. This would appear to be difficult in Wayland.

    yes, indeed, your memory serves right. that has been brought up on
    emacs-devel (in fact, I have some of those threads saved so that I can
    go back when I get some hacking time). in truth, I think the
    implementation in Emacs likely won't be (fully) reusable, since wayland
    will likely never let clients self-position like Emacs wants to.

    it has been said to me that a complete Emacs port 'requires' arbitrary positioning and cursor reading (IIRC, probably other stuff I'm
    forgetting). I'm afraid that's likely not to happen (but I'm not too
    sad about losing that functionality either).

    You say "yes, intentionally, wayland is declarative wrt window
    positioning". What does that mean when you replace abstract words like "declarative" with concrete sentences? What is declaring what to what
    else in this context, and what does that have to do with not being able
    to position windows?

    well, it means that there will be a long and arduous process of
    standardization and development to most things people conventionally use
    these 'hacks' in X for, including this one:

    https://gitlab.freedesktop.org/wayland/wayland-protocols/-/merge_requests/18

    I haven't read the full proposed protocol, but I suspect it is the right
    one here.

    the disadvantage of this approach is that it's slow as molasses, the
    advantage is that there will be a descriptive way to do something with
    no room for 'whoops, forgot this workaround' or 'huh, forgot that
    property', etc, and slightly different behaviour from program to
    program.

    Later on, you say "where /relative to a surface/". I think "surface" is
    a word with particular meaning in Wayland, and using it in a Gentoo list without explanation is less than helpful. What does "surface" mean in
    this context? Is the entire screen such a "surface"?

    well, the term is somewhat abstract because it is an abstraction :-)

    a surface is a rectangle of a program with some pixels on it that can
    receive events and stuff (so, a window of some kind). a toplevel
    surface is what we conventionally think of as proper windows (like the
    one I'm typing this email in, which Emacs calls a frame); a popup
    surface is for, well, popups (C-<right click> menu for instance).

    respectively, they are defined as:

    https://wayland.app/protocols/wayland#wl_surface
    https://wayland.app/protocols/xdg-shell#xdg_toplevel
    https://wayland.app/protocols/xdg-shell#xdg_popup

    the interface responsible for positioning popups is:

    https://wayland.app/protocols/xdg-shell#xdg_positioner

    the latter allows the client to tell the server where to put a new
    surface compared to the current one. this information is later passed
    to the server with a request to create a popup surface.

    So, is it possible in Wayland to record a configuration of windows,
    their sizes and positions, then restore these on starting a program
    again? If not, that would appear to be a design bug in Wayland. What
    am I missing?

    well, more like a todo or wip (which, yes, I'm aware is a frequently
    cited criticism, but I promise it's not as bad as it seems - as
    mentioned, I very rarely run into real limitations). wayland is
    extensible, so even though it is seeing relatively wide adoption, it
    also sees active development.

    (joke) but surely this is entirely unnecessary, because who ever closes
    emacs? :)

    have a lovely evening
    --
    Arsen Arsenović

    --=-=-Content-Type: application/pgp-signature; name="signature.asc"

    -----BEGIN PGP SIGNATURE-----

    iIcEARYKAC8WIQT+4rPRE/wAoxYtYGFSwpQwHqLEkwUCZrKfUBEcYXJzZW5AZ2Vu dG9vLm9yZwAKCRBSwpQwHqLEkydGAP4sjlzxbrmXXZwc1ENPTyRSZr11HJFiEd1+ viF7lvyb4QEAjTem27NOvO+GN/MsfBVL4vLfiBuvGRB5sWGBhnDQEgs�yl
    -----END PGP SIGNATURE-----

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Matt Jolly@21:1/5 to Wols Lists on Wed Aug 7 11:10:01 2024
    Hi,

    On 6/8/24 02:58, Wols Lists wrote:

    Last I investigated, sddm had a *hard* dependency on X11. So even if
    you're running a Wayland system (like I am) you need X installed so that
    sddm will work.


    That's not quite correct; it's been possible to run SDDM directly as a
    Wayland session for quite a while. I have several machines that use `kwin_wayland`: https://wiki.gentoo.org/wiki/SDDM#Kwin

    Since Plasma 6 I haven't noticed any issues, it used to be a _little_
    weird for Plasma 5 if you (e.g.) logged out.

    Cheers,

    Matt

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Michael@21:1/5 to All on Wed Aug 7 11:11:35 2024
    On Wednesday, 7 August 2024 10:04:28 BST Matt Jolly wrote:
    Hi,

    On 6/8/24 02:58, Wols Lists wrote:
    Last I investigated, sddm had a *hard* dependency on X11. So even if

    you're running a Wayland system (like I am) you need X installed so that
    sddm will work.


    That's not quite correct; it's been possible to run SDDM directly as a Wayland session for quite a while. I have several machines that use `kwin_wayland`: https://wiki.gentoo.org/wiki/SDDM#Kwin

    Since Plasma 6 I haven't noticed any issues, it used to be a _little_
    weird for Plasma 5 if you (e.g.) logged out.

    Cheers,

    Matt

    Is this on an openrc or systemd installation? Does SDDM launch on VT1?


    -----BEGIN PGP SIGNATURE-----

    iQIzBAABCAAdFiEEXqhvaVh2ERicA8Ceseqq9sKVZxkFAmazSFcACgkQseqq9sKV ZxlAcRAAimbDywfyeCP3PC9fjps5fmMmjwLBKkC2nmr42j0YRvItww2aiZbs1VYz p2dZSbJoVvViKywKmiskoCcrlIZ05gfU/PNnczkg9mJO3aEBRMk8+gS/N+Dc+0rH QsuQSN2WPiEcjU4bGl/Iqi/xDGLBXd1gVYplsgUTbmxW4u7Vm/QtrpdAMh6Mx6Dk 0qXvAF3/Liq99VTeWboiSp4MESihtyaHAf2G55qhpqm5gGhrl35mkj/FWKkWX18A x7iSEJejTNF/DrEqD9g+Z9qkZUlVSnIma79AdghRd/p+UrMFxGNk+cvqBkqCjXt6 0MHOu1i8eCDOB4OHoUs1vHjsfrUF6orMR2LzcekkIcsnknZYLtbcHWBzRFJnW8aI ZiqR3C9tCtyBtVH2kf1iqlE8wZUx1Ld+mEvge55JBKmq+GOp9TAPXNPDpDGjGD4Y TB8Wny4EEg7mRInubgTZhhHOzHMSfWD+Os4Zv559Lbbk1gLG6Ts0uCSb7qzZsyhc qWnQOT/kq4dAVxp+LW9Ie3n70TwiTTPCqgh5aa16l4zslu3MPLW9w44bs6wazgZH E5ZmLXTJ6/GjdQxXikVZ94+fb/fZkekDj5LpmuNfseUWnPciRf0Vb/03cW+VR05I theljO+CK4Bf2pVSVn+lrrxSpuZcl86nJdg8QnbjCms/ZOqv3W4=
    =iU/Q
    -----END PGP SIGNATURE-----

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Alan Mackenzie@21:1/5 to Wol on Wed Aug 7 16:00:02 2024
    Hello, Wol.

    On Tue, Aug 06, 2024 at 23:08:42 +0100, Wol wrote:
    On 06/08/2024 19:31, Alan Mackenzie wrote:
    So, is it possible in Wayland to record a configuration of windows,
    their sizes and positions, then restore these on starting a program
    again? If not, that would appear to be a design bug in Wayland. What
    am I missing?

    That - unlike X - is because windows cannot say where they are going to
    go. They can *ask* where they want to go, which isn't the same thing.

    How does it differ in practice? Under what circumstances would a
    request to display a window at a particular place result in it being
    displayed somewhere else?

    Iirc, X behaves like Windows, which means applications can *seize*
    focus, which drives me up the wall on occasion at work. I'll have an
    Excel macro running, which takes maybe 3 or 4 minutes. So I go into
    let's say Slack. Excel triggers something (google drive?) which grabs
    focus and disappears, so all of a sudden I *think* I'm gaily typing into Slack. But focus has been stolen and I'm typing into a vacuum -
    EXTREMELY frustrating, especially as I don't actually know what's going on.

    I don't understand what these issues with focus have to do with
    positioning a window. Though I can appreciate them causing problems.
    There would appear to be a clash between Wayland running within a
    GNU/Linux running as a Windows subsystem, and the Windows itself -
    presumably Windows allows a Windows application to steal focus from a
    Wayland application in this situation.

    In Wayland, you can't steal focus. But as a side effect, it's Wayland
    that controls the window, not the application. So Wayland is more
    secure, but that comes with unavoidable side effects that you don't like...

    How does Wayland controling the Window lead to an application program's inability to position it? I can't see the connection.

    Just as a bit of context; I've not yet tried Wayland, and for most of my
    work (including Emacs) use a Linux console.

    Cheers,
    Wol

    --
    Alan Mackenzie (Nuremberg, Germany).

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Wol@21:1/5 to Alan Mackenzie on Wed Aug 7 20:40:01 2024
    On 07/08/2024 14:53, Alan Mackenzie wrote:
    Hello, Wol.

    On Tue, Aug 06, 2024 at 23:08:42 +0100, Wol wrote:
    On 06/08/2024 19:31, Alan Mackenzie wrote:
    So, is it possible in Wayland to record a configuration of windows,
    their sizes and positions, then restore these on starting a program
    again? If not, that would appear to be a design bug in Wayland. What
    am I missing?

    That - unlike X - is because windows cannot say where they are going to
    go. They can *ask* where they want to go, which isn't the same thing.

    How does it differ in practice? Under what circumstances would a
    request to display a window at a particular place result in it being displayed somewhere else?

    It probably doesn't make much difference most of the time. Where it particularly makes a difference is, I believe, modal windows, and
    windows which demand to be placed "on top".

    In Windows or X, a window which demands to be placed on top of
    everything else will get what it asks for. And it doesn't give a monkeys
    about what the user is actually doing at the time.

    With Wayland, if the user says "I want this window on top", then the application can't over-ride it - Wayland will force it to pop-under.

    Iirc, X behaves like Windows, which means applications can *seize*
    focus, which drives me up the wall on occasion at work. I'll have an
    Excel macro running, which takes maybe 3 or 4 minutes. So I go into
    let's say Slack. Excel triggers something (google drive?) which grabs
    focus and disappears, so all of a sudden I *think* I'm gaily typing into
    Slack. But focus has been stolen and I'm typing into a vacuum -
    EXTREMELY frustrating, especially as I don't actually know what's going on.

    I don't understand what these issues with focus have to do with
    positioning a window. Though I can appreciate them causing problems.
    There would appear to be a clash between Wayland running within a
    GNU/Linux running as a Windows subsystem, and the Windows itself -
    presumably Windows allows a Windows application to steal focus from a
    Wayland application in this situation.

    In Wayland, you can't steal focus. But as a side effect, it's Wayland
    that controls the window, not the application. So Wayland is more
    secure, but that comes with unavoidable side effects that you don't like...

    How does Wayland controling the Window lead to an application program's inability to position it? I can't see the connection.

    Because Wayland (or rather the user, through Wayland) DICTATES what the application is allowed to do.

    In practice it may not make much difference. But the user CAN tell
    Wayland "this space is reserved for AppA", and if AppB comes along and
    says "I want to put my window over AppA", Wayland will tell it to bugger
    off.

    It's basically Wayland's security stance - if I the user *think* I am interacting with AppA, Wayland blocks AppB from taking over and tricking
    me into eg leaking my password. All you need is an app with a hidden
    window that seizes focus when it sees a password pop-up appear, and your secrets are leaked ...

    Just as a bit of context; I've not yet tried Wayland, and for most of my
    work (including Emacs) use a Linux console.

    I've tried to switch to Wayland, but as you can tell I haven't
    necessarily managed it...

    Cheers,
    Wol

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)