Getting KDE onto commercial hardware

It's a long shot, but Linux really needs a success story in graphical consumer devices outside Android. I think that the potential is there, one moderately successful Linux-based device business can set the ball rolling for many more and from then on, there would be no stopping it.

The state of Linux desktop is that things mostly work, but it's not rare to experience subtle driver- and display stack bugs. Due to lack of resources, bug reports often languish, making it a bad time investment to even report such bugs. Whereas if these bugs were critical to a business that's selling devices, they would find the means to improve the situation.

One such success story is Valve's continued investment in improving the Linux 3D graphics stack and Steam experience on Linux. (Well, success story for Linux at least, not sure about Valve's business). Besides their direct involvement, I think they also play a significant role in AMD's recent renewed interest in upstreamed open source Linux drivers and probably other knock-on effects. The end result is that gaming on Linux went from a horrible experience a few years ago, to a viable gaming OS.

Valve's Steambox hardware, unfortunately has not panned out so far. Similar to so many other attempts, like Ubuntu for tablets, Openmoko and probably many more. Dell's preinstalled Linux offerings, System76, Purism etc aren't faliures by any means, but have not brought mainstream attention to the platform.

My thoughts exactly. A distro is not just the graphical environment: many many things need to be shipped to the user (the kernel, libc, ssh, a thousand other little programs), and many things need to be maintained on the distro side (a wiki or some other help forum, a bug tracker, a package tracker, some build infra, some CI infra, distribution server and CDN or mirrors). So if you take the route of starting your own distribution seriously, you need to put an *enormous* amount of up-front work before anything is even shipped to users, and then a lot on-going maintenance just to keep things ticking.

There is the obvious alternative: create a spin as part of some $distro umbrella. Reuse as much as can be reused, lean on the existing build, test, qa, and distribution infrastructure. Use your time for the parts where you are the expert and can make the biggest difference.

I don't know too much about Ubuntu and Neon, but I can speak about Fedora: we would *love* to have more upstream KDE contributors participating in the KDE spin. KDE is not the primary desktop so it doesn't get as much attention as Gnome. Maintainers for KDE are stretched thin and kde-specific bugs are only slowly handled. But what is packaged in Fedora is generally just the latest version that upstream provides and there is no reason why it couldn't be used to showcase upstream work.