RFC: xserver update strategy in F21+

[Adam Jackson]

On Tue, 2014-11-18 at 08:24 +0100, Thorsten Leemhuis wrote:

> But in the end Fedora and its kernel maintainers didn't care. Which
> might be the right thing to do for X as well, because companies then
> learn that they need to keep track of ongoing development and users
> notice some of the risks these driver bear.

Holding free drivers hostage to closed ones is obviously off the table;
if I need to update X to get a feature into Fedora's radeon driver, and
catalyst hasn't caught up yet, catalyst is going to lose that fight.

But even the best free operating system is worthless if you can't use
it, and the binary drivers do have actual features beyond the free ones
that people do in fact depend on to get work done.  So, while we
continue to work on getting to feature parity, I don't want to be
arbitrarily harsh to people who don't have a choice in the matter.

It sounds like people are reasonably comfortable with a kernel-like
update policy.  If we run into that kind of ABI scenario with binary
drivers, presumably the best course of action is to bring it up for
discussion case by case and refer it to FESCO if it's especially
contentious.

Meanwhile, I've started assembling a crib sheet:

https://fedoraproject.org/wiki/RebasingXserver

I'll flesh that out more as we approach the first actual rebase like
this, which if I had to guess would likely be early 2015.

- ajax