SUMMARY I can setup keyboard, mouse, monitors, audio I/O, printers, joysticks, tablets ... Everything but the cameras. STEPS TO REPRODUCE 1. Open system settings 2. search for "cam" OBSERVED RESULT No items matching my search EXPECTED RESULT At least a basic item to setup my PC cameras. SOFTWARE/OS VERSIONS Windows: n/a macOS: n/a Plasma Version : 5.27.10-1 KDE Version : 23.08.4-1 Frameworks Version : 5.113.0-1 Qt5 Version : 5.15.11+kde+r12-1 Qt6 Version : 6.6.1-3 ADDITIONAL INFORMATION Not really a bug. More a absent-mindedness. Cameras are here since 20+ years.
Out of curiosity, what would you expect to configure there?
Resolution. Mirroring mode. Rotation. Brightness. Whatever you can adjust.
How would apps read those settings, though? Typically they just open up the raw device right now, which means any settings we make around this would be ignored and useless. What we would have to do is write a wrapper library that facilitates interacting with a camera which does read our settings. But then the problem is that every app that uses the camera would have to be ported to use that library. This is feasible for KDE apps, but not for 3rd-party apps, who it would be difficult to impossible to convince to use a KDE-specific library for accessing the camera. As a result it would be a "broken promise" global setting of the type that we are trying to move away from; see https://community.kde.org/Get_Involved/Design/Frequently_Discussed_Topics#%22Broken_promise%22_global_options. What would be a more appropriate approach would be to write the library upstream in the XDG/FreeDesktop namespace, (like Libinput), define the settings with a cross-desktop spec, and then begin the painful process of convincing everyone in the world to use that library. This would probably take at least 10 years, and that's if the resources for such a project were even to materialize. The reason this worked for Libinput is because Red Hat paid Peter Hutterer a full-time salary to consistently put in that effort for over 10 years. Prevailing wages being what they are, I'd guess that Red Hat has sunk more than $1 million into the project. As you can see, this would be an enormous amount of work. It's a good idea, but it would be a multi-year project and I'm afraid is rather outside of the scope of a KDE bug report. :)