From owner-freebsd-hackers Tue Jan 16 15:04:23 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id PAA24441 for hackers-outgoing; Tue, 16 Jan 1996 15:04:23 -0800 (PST) Received: from sivka.carrier.kiev.ua (root@sivka.carrier.kiev.ua [193.125.68.130]) by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id PAA24253 for ; Tue, 16 Jan 1996 15:03:49 -0800 (PST) Received: from elvisti.kiev.ua (uucp@localhost) by sivka.carrier.kiev.ua (Sendmail 8.who.cares/5) with UUCP id AAA23704 for hackers@freebsd.org; Wed, 17 Jan 1996 00:58:53 +0200 Received: from office.elvisti.kiev.ua (office.elvisti.kiev.ua [193.125.28.33]) by spider2.elvisti.kiev.ua (8.6.12/8.ElVisti) with ESMTP id AAA15426 for ; Wed, 17 Jan 1996 00:36:39 +0200 Received: (from stesin@localhost) by office.elvisti.kiev.ua (8.6.12/8.ElVisti) id AAA05692; Wed, 17 Jan 1996 00:36:37 +0200 From: "Andrew V. Stesin" Message-Id: <199601162236.AAA05692@office.elvisti.kiev.ua> Subject: Re: user management stuff To: jkh@time.cdrom.com (Jordan K. Hubbard) Date: Wed, 17 Jan 1996 00:36:36 +0200 (EET) Cc: stesin@elvisti.kiev.ua, hackers@freebsd.org In-Reply-To: <11329.821821604@time.cdrom.com> from "Jordan K. Hubbard" at Jan 16, 96 11:46:44 am X-Mailer: ELM [version 2.4 PL24alpha5] Content-Type: text Sender: owner-hackers@freebsd.org Precedence: bulk Comments for review: # std options for all invocations: # [-uid uid] [-gid gid] [-user user] [-group group] [-home homedir] [-name name] # [-shell shell] [-realhome realhome] [-after uid] # 1. What about a one-char options, i.e. "-u" == "-uid" ? # -uid specify the uid to use, otherwise it will be chosen automatically # (see also -after). # # -gid specify the gid to use, otherwise it's set automatically to track # the uid. # # -user Specify the username to use [no default] 2. Maybe better to have login name not as an option, but as a required (single) argument? # -group Check that gid also exists under name group - if not, it should # be added under this name (e.g. group:*:gid:user). Default is not # to mess with /etc/group. 3. Considering a Current Policy (separate gid for each uid, (c) Rodney Grimes; I personally like this approach) how can I skip /etc/group modification? # -home Specify the location of the home directory. By default it will # go to /home/${user} if not overriden in /etc/adduser.cf ^^^^^^^^^^ 4. "Are you sure in is wise?" :) I mean the fact that this file exists at all. # -name Specify full name (+ any other GCOS info). 5. You mean "a single string arg, possibly with a comma separated list of GCOS fields in it"? # -shell Specify a different shell (than the default) to use. # # -realhome If different from home, make home dir here instead and # make a symlink from home to here. Nice approach. # -after If chosing uid automatically, start search after this uid. # Search will stop at the next free "hole". # # # Commands: # # adduser -add [..stdopts] [-skel skeldir] Got it. # adduser -delete [key] 6. Is "key" == "logname" or it may be specified as, for example, "-uid XXX" or "-realname 'Random J. User'" ? # adduser -exists [key] 7. To be able to get an exit status of 0 if "yes" 1 otherwise? # adduser -modify -uid 701 -name "Irving Q. Steenbottle" # # Finds uid 701 and changes the name field accordingly, leaving # other fields unaltered. 8. This means priorities of matches? I.e. you doesn't want to modify uid if mr. "Irving Q. Steenbottle" is there already with uid 700? # > What the program _must not_ do? What kind of UI is # # Crash? Ruin the password file? :-) (-; ... or change spwd.db without altering master.passwd. # I think the UI issue should be left alone to implement one level # further up, if desired. The first UI interface should probably be # libdialog based, just for consistency. Ok. Leaving it alone for now. # Oh yeah, it should also read an /etc/adduser.cf for default # configuration information (which I think it should *not* prompt the # user about - just choose reasonable defaults! The "adduser # configuration" phase on the first startup of our current adduser # really confuses a lot of people! :-( Oh yes :-) See q. 4 above. I'm saving your original draft spec and waiting one day before doing actual start (anyway, tomorrow I'll be busy lurking aroung looking how to replace a Conner CPF1060S where news spool lived until an hour ago :( It finally died, after 3 days of NCR writing error msgs to the console.) -- With best regards -- Andrew Stesin. +380 (44) 2760188 +380 (44) 2713457 +380 (44) 2713560 An undocumented feature is a coding error.