Manual Pages  — CHPASS

The gid field is the group that the user will be placed in at login. Since BSD supports multiple groups (see groups(1)) this field currently has little special meaning. This field may be filled in with either a number or a group name (see group(5)).

The class field references class descriptions in /etc/login.conf and is typically used to initialize the user's system resource limits when they login.

The change field is the date by which the password must be changed.

The expire field is the date on which the account expires.

Both the change and expire fields should be entered in the form "month day year" where month is the month name (the first three characters are sufficient), day is the day of the month, and year is the year.

Five fields are available for storing the user's full name , office location, work and home telephone numbers and finally other information which is a single comma delimited string to represent any additional gecos fields (typically used for site specific user information). Note that finger(1) will display the office location and office phone together under the heading Officeamp;:.

The user's home directory is the full Unix path name where the user will be placed at login.

The shell field is the command interpreter the user prefers. If the shell field is empty, the Bourne shell, /bin/sh, is assumed. When altering a login shell, and not the super-user, the user may not change from a non-standard shell or to a non-standard shell. Non-standard is defined as a shell not found in /etc/shells.

Once the information has been verified, chpass uses pwd_mkdb(8) to update the user database.

See pwd_mkdb(8) for an explanation of the impact of setting the PW_SCAN_BIG_IDS environment variable.

Note: these exceptions only apply when the NIS master server is a FreeBSD system).

Consequently, except where noted, the following restrictions apply when chpass is used with NIS:

1. Only the shell and GECOS information may be changed. All other fields are restricted, even when chpass is invoked by the super-user. While support for changing other fields could be added, this would lead to compatibility problems with other NIS-capable systems. Even though the super-user may supply data for other fields while editing an entry, the extra information (other than the password — see below) will be silently discarded.

   Exception: the super-user on the NIS master server is permitted to change any field.

2. Password authentication is required. The chpass utility will prompt for the user's NIS password before effecting any changes. If the password is invalid, all changes will be discarded.

   Exception: the super-user on the NIS master server is allowed to submit changes without supplying a password. (The super-user may choose to turn off this feature using the -o flag, described below.)

3. Adding new records to the local password database is discouraged. The chpass utility will allow the administrator to add new records to the local password database while NIS is enabled, but this can lead to some confusion since the new records are appended to the end of the master password file, usually after the special NIS '+' entries. The administrator should use vipw(8) to modify the local password file when NIS is running.

   The super-user on the NIS master server is permitted to add new records to the NIS password maps, provided the rpc.yppasswdd(8) server has been started with the -a flag to permitted additions (it refuses them by default). The chpass utility tries to update the local password database by default; to update the NIS maps instead, invoke chpass with the -y flag.

4. Password changes are not permitted. Users should use passwd(1) or yppasswd(1) to change their NIS passwords. The super-user is allowed to specify a new password (even though the "Password:" field does not show up in the editor template, the super-user may add it back by hand), but even the super-user must supply the user's original password otherwise rpc.yppasswdd(8) will refuse to update the NIS maps.

   Exception: the super-user on the NIS master server is permitted to change a user's NIS password with chpass.

There are also a few extra option flags that are available when chpass is compiled with NIS support:

-d domain
	Specify a particular NIS domain. The chpass utility uses the system domain name by default, as set by the domainname(1) utility. The -d option can be used to override a default, or to specify a domain when the system domain name is not set.
-h host
	Specify the name or address of an NIS server to query. Normally, chpass will communicate with the NIS master host specified in the master.passwd or passwd maps. On hosts that have not been configured as NIS clients, there is no way for the program to determine this information unless the user provides the hostname of a server. Note that the specified hostname need not be that of the NIS master server; the name of any server, master or slave, in a given NIS domain will do. When using the -d option, the hostname defaults to "localhost". The -h option can be used in conjunction with the -d option, in which case the user-specified hostname will override the default.
-l	Force chpass to modify the local copy of a user's password information in the event that a user exists in both the local and NIS databases.
-o	Force the use of RPC-based updates when communicating with rpc.yppasswdd(8) ("old-mode"). When invoked by the super-user on the NIS master server, chpass allows unrestricted changes to the NIS passwd maps using dedicated, non-RPC-based mechanism (in this case, a Unix domain socket). The -o flag can be used to force chpass to use the standard update mechanism instead. This option is provided mainly for testing purposes.
-y	Opposite effect of -l. This flag is largely redundant since chpass operates on NIS entries by default if NIS is enabled.

Robert Morris, Ken Thompson, UNIX Password security,