Darren Reed writes:

>In some mail I received from Jonathan Stone, sie wrote
>> Maybe breaking AFS server compatibility is the right thing, but we
>> should be aware we're doing that, and/or make it an option.

>Well, my hope would be, that if there is some sort of finite usage of
>those bits that an appropriate (set of) bit pattern(s) could be somehow
>"reserved" for AFS compatibility.  If arbitary values go there, that
>would sort of void this idea.

I don't know, but AFS does have ACLs, and I imagine AFS severs use
these bits for the same purpose which started this discussion:
a reference to an inode containing per-file ACL info.

Someone with access to AFS source should really comment.