Simply put, I would like to be able to run a headless, persistent, detachable/reconnectable plasma-wayland session with a VNC server, as wlroots-based compositors are able to do with the aid of WayVNC[1]. I am aware that kwin_wayland has a virtual backend, which was mooted when introduced as a means to provide such functionality[2] - that was a long time ago so I do not know whether that would still be considered the best starting point. I am also unsure how easily reproducible is what wlroots provides that WayVNC needs, or whether it would be necessary to implement everything including the server in a "KDE Way"; so I don't know if this will be a big ask or a VERY big ask. I would just like to know if there's an avenue and any interest towards it. I think it could be very beneficial in cloud systems, VMs and even plain old LAN usage once X11 finally shuffles off. [1] https://github.com/any1/wayvnc [2] http://blog.martin-graesslin.com/blog/2015/10/september-update-for-plasmas-wayland-porting/
It sounds like you rather need a way to tell the plasma start script to start kwin_wayland with the virtual backend and also provide a way to pass additional arguments to configure the mode size or the refresh rate.
(In reply to Vlad Zahorodnii from comment #1) > It sounds like you rather need a way to tell the plasma start script to > start kwin_wayland with the virtual backend and also provide a way to pass > additional arguments to configure the mode size or the refresh rate. *If* the virtual backend is viable for such a purpose (this is one aspect on which I wasn't able to get up-to-date info), then what you detail would be a necessary part of the puzzle, yes. But there is also how inputs are handled. I don't understand very well how responsibilities divide up between what the compositor needs to do to enable it and what a VNC server, as a distinct entity, needs to implement for itself.
VNC support in kwin is not something we intend to do directly, we support 3rd party apps streaming content (see krdb or others) Having a mechanism to specify the virtual backend from startplasma makes sense.
Thank you for clarifying things for me. So it sounds like the virtual backend *can* provide what a 3rd-party VNC server would need on the output side. I will look elsewhere for what would need doing on the input side. Can I just enquire regarding virtual backend outputs, can they be resized dynamically? And could a virtual output be added (dynamically or at launch) to a "bare metal" session as a second screen?