From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Thu, 20 Jul 2000 16:56:19 +0200 From: Lucio De Re To: 9fans mailing list <9fans@cse.psu.edu> Subject: Re: [9fans] UID and GID renumbering Message-ID: <20000720165619.P3081@cackle.proxima.alt.za> References: <200007201443.KAA00115@smtp1.fas.harvard.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <200007201443.KAA00115@smtp1.fas.harvard.edu>; from Russ Cox on Thu, Jul 20, 2000 at 10:43:53AM -0400 Topicbox-Message-UUID: e7d4077a-eac8-11e9-9e20-41e7f4b1d025 On Thu, Jul 20, 2000 at 10:43:53AM -0400, Russ Cox wrote: > > Just how dangerous is it to renumber the entries in /adm/users? > I need to know what kind of crises I may trigger, broadly. > > The numbers in /adm/users are used in the > on-disk structures in both kfs and the real > file server. Unless you're going to rewrite > the file system, you probably need to leave it as is. > Let me ask a different question, then: I installed the 3ed distribution effectively using the existing 2ed /adm/users, with the result that /usr/glenda is reported as owned by adm. I think this is not serious, but what I can't quite pinpoint is which of upas=10001 and upas=10003 (old /adm/users) is the case. My proxima replacement for bootes is another candidate for straightening out, it was =2 in the 2ed system, and it's not even in the /adm/users file in the 3ed distribution, understandably, since the file is not used by the file server. ++L