Hi Kevin, > What might be a way forwards here, assuming you want to keep the old files > around, would be to change the group-name for files that become orphaned > through a package upgrade to have a versioned group name.
That would get on my nerves. It just isn't "neat" enough :-) > As to the group names, > >> The original hint does not provide much guidance for what group name to >> assign to each package > > The original hint gives THREE choices of approach ! You are correct, but they all have problems. If I tell you that you can either shoot yourself in the foot, or stick your hand in a vat of acid, or jump off a cliff, I'm not really giving you a pleasant course of action to follow, even though I've generously offered three choices. The hint's first choice (group name = user name) isn't really using the group information **for** anything. The group information in this case is redundant and unhelpful. The advantage regarding setuid that you point out is not annulled by the scheme I've suggested. Regarding the second choice, it appears to me that the selection of groups is rather ad hoc. Even if one customises the selection instead of following the appendix, it is still ad hoc. I have trouble seeing how such a schema could possibly fit all possible packages that I may wish to install now and in the future. The third choice, as pointed out in the hint, is organisation-dependent, and outside the scope of the hint. Anyway, it really only seems like a variant of the second choice, but we now insist on a correspondence between groups inside the computer and meat outside the computer. My proposal, I think, sort of hybridises the first two choices. I believe it takes advantage of the good points of both, and does not share the down-sides. Cheers, Tim -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
