On Sat, Aug 25, 2001 at 09:36:37PM -0400, Ken Cross wrote:
> How it's stored on disk is implementation-dependent and varies considerably.

Which, it seems to me, is the only remaining question bearing debate
on tech-kern.

FFS does not seem to have bits free for this. Certainly not for a
plausibly infinite block of metadata (well, okay, you couldn't have
more than 65536*2 sets of credentials, but that's a lot of bits).
And (imho) it belongs at the UFS layer anyway, so that we get it
in other file systems. (Yes, I have a specific LFS disk in mind
where I want ACLs, thanks for not asking.)

So, then, what can we do to allow an ACL-aware kernel to still
behave correctly with an old-style {F,U}FS partition (that's the
easy part), and what can we do to make a non-ACL-aware kernel able
to read a file system with ACLs (that's the hard part).

In my opinion, there's no need for the second part. As long as
install media sticks to the old style (and why shouldn't it? Look at
Apple and HFS versus HFS+ in MacOS 8 and 9), the only time I'm
going to pop a disk with ACLs into a machine that doesn't have them
is when I'm doing some kind of data recovery. And I can deal with
building a kernel for that. (Well, actually, I'll probably never
stick a disk with ACLs into a machine without, as I'll probably use
them everywhere, as I know how much better a system they are having
used them under Solaris. But whatever.)

What's really lacking is anyone with the knowledge, time, and
interest to make this happen. I'm short at least one of those at
present myself.

-- 
       ~ g r @ eclipsed.net