09:18 <diederik> I'm not part of the LE team and nowhere near as knowledgeable as chewitt, but I'm pretty sure that including BSP code makes it technically unfeasible and possibly even technically impossible

09:19 <diederik> In general: BSPs are each their entirely different code bases, often based on a (very) old kernel version and then heavily modified to make things 'work' for that particular device

09:21 <diederik> And they don't take into consideration (any) other devices. So it isn't just 'grabbing' a few patches. You'd have completely different kernel trees for any non-mainlined devices

09:23 <diederik> So that would be a HUGE maintenance burden in itself and then you also have the issue that the kernel interfaces that kodi/LE uses are different, which makes that extremely hard as well.

09:24 <diederik> Long story short: Your best bet is getting HW support in the upstream Linux kernel, thus upstream (HW) device support

11:50 <tito-me-doe> LibreELEC’s team would not nor should not need to worry about BSPs themselves. Neither working with them nor supporting them. The idea behind BootELEC is really about moving device support into a larger project with more organizations maintaining it. It’s about separating the parts that give firmware projects their unique identities from the boring stuff that’s just the minimum requirements to get Linux running on X

11:50 <tito-me-doe> device. And while I agree mainline support is always the goal, it takes time , and the kernel isn’t the only piece of supporting a device

11:51 <tito-me-doe> Ideally, a project like this would only enable more eyes and hands on making mainline patches

11:53 <tito-me-doe> But each organization maintaining project BootELEC would only need to worry about whatever devices/kernels they intend to maintain. They are not forced to use nor contribute directly to BSP code