[zaxcellent]

There is a lot of detail about this in the kernel docs: https://www.kernel.org/doc/html/latest/gpu/drm-uapi.html#ope...

The gist of it is that breakage is OK if the userspace isn't open-source and that new uAPIs come around every few years due to the rapid pace development of GPUs.

I worked on Chrome OS graphics for years on and one of my earliest projects was to bring DRM/KMS (then a newish interface) to the Cirrus display card (an ancient card that QEMU happens to emulate).

[michaelmrose]

Does it specifically mention the kernel breaking X? One would think that would be the topic of discussion years before anyone actually did it.

[zaxcellent]

I don't think the kernel guarantees that if X runs against Linux today, then it will run against the same X on new kernels on new hardware in perpetuity. It's more along the lines of, "if I one calls the read syscall with parameters a, b, c, it won't suddenly require a 4th parameter, or demand they be in a, c, b order."

Anyways, if the kernel developers decided they wanted to invent a brand new uAPI that does graphics, I wouldn't at all be surprised if new GPU drivers only implemented the new uAPI if their target userspace programs all use the new uAPI instead of an older uAPI. Nothing in userspace is "broken" because it will seem to the old uAPI using userspace as if there is no driver for the new GPU.

[michaelmrose]

To be clear in the Wayland fetishists fantasies they wouldn't need to convince people to all switch to wayland circumstances would foreclose on any further debate in reality all players have an incentive to keep existing software working.

If one calls the _ syscall with parameters a b and c it won't suddenly require a 4th or demand they be in a c b order is literally what makes X11 continue working against newer kernels. Providing new functionality wherein if you don't call foo before calling _ with a b and c it fails to perform its function would be functionally the same as changing the order of parameters it would be breaking user space.

I don't see how you could possibly lawyer yourself past that, I don't see that the adults in the room who have to support customers in rhel 8 through 2029 can do anything but keep doing minimum work to keep x and the kernel playing nicely. I don't see how nvidia is going to go from supporting hardware for 10-18 years even on niche hardware to dropping support for X so quickly. I also don't see how you can meaningfully talk about the kernel dropping support for X without basing your discussion on actual kernel developers who logically would talk about kernel foo will be the last to support X probably years before it actually happens. I'm totally sure that python 2 -> 3 took over a decade after it was a completely suitable replacement but we will totally manage to replace X before its replacement is fully baked and the kinks worked out.

In short the whole premise is completely and totally premature wishful thinking by people who want their own way. I'll pour one out for X when its actually and in fact dead.