Thus spake Frederick Bruckman ("FB> ") sometime Today...

FB> Of all the proposals, the only one that improves on the current
FB> one (IMO), is to *not* bump uname, and let LKM's get their needed
FB> information some other way.

Good grief, is this _still_ just about LKMs?!?  You'd think that the
fate of the world rested on this!

How about some way of presenting a concrete way (to carbon-based
lifeforms) trace back to "the last known good version that worked" without
having to by-guess-and-by-golly or go digging through mail archives?  One
would think that the ufs2 changes might have taught us _some_thing in this
regard!

If you're worried about lkms, create a 'kern.branchdate' or something
which would provide appropriate information.

If we're still version-changing based solely on what LKMs want and no
other (life-impacting, if you will) factor, I would say that our
idea of versioning is SORELY misguided.

				--*greywolf;
--
NetBSD: Unix With Balls.