# $Header: /usr/src/nidentd-2.0/RCS/README,v 2.0 92/05/04 17:42:49 nigelm Exp $
New Ident Daemon - nidentd
~~~~~~~~~~~~~~~~~~~~~~~~~~
Licensing Information
---------------------
This program is released into the public domain and can be used by
anyone who wants to. Vendors may include it into their distributions
if they want to without any restrictions.
Although there are no formal restrictions on use, modification and
distribution, I would appreciate it if the following guidelines were
adhered to just to help me keep sane!
+ If the program is distributed in a modified form, the version
numbers are changed to make it distinct from versions which I
distribute. (This means that I know when I get questions about
a derivative version as opposed to one I know something about).
+ Please inform me (as a courtesy not a requirement) if you distribute
a modified version, and send the mods back to me.
+ Please leave my name and the reference to the Univerity of York
on any derived version *under* any reference to you as the
supporting author (ie you support it!!).
Program Information
-------------------
This is a program that implements the RFC931 identification server. It
is a direct modification of Peter Eriksson's (pen@lysator.liu.se) pauthd
to handle mach and System V derived Unix variants. Most of the code and
even more of the documentation is very similar to Peter's version, which
was very much inspired by Dan Bernstein's original 'authd'. The additional
functionality of Dan's authd (which had command line functions to trace a
TCP/IP connection) has also been rolled into the code.
Unlike authd, this program does not use 'netstat' to get some of the
information) It uses the kernel information directly.
Modifications have been made to bring this code into line with the new
RFC for the identity server (which is why this called pidentd rather than
some authd related name).
The daemon currently *only* contains code for a NeXT system (tested under
NeXTstep 2.0) and a MIPs system. Earlier development versions carried
over code from Peter's version for other types of computer systems, but
I have removed that for now since I cannot test it on other platforms.
I intend to add functionality for an SGI system, but this is a major
challenge since the kernel file structure is so different.
Please let me know if you find any bugs, or have ported it to other
machines (and care to share the changes with me and the world!).
There are a few options to the daemon that may help you debug it if
porting it to new machines, -V prints the version stamp and -d enables
debug messages (prints some informative error instead of "UNKNOWN ERROR").
NOTE: One should NOT use the -d option when using it for normal use!
Other Files
-----------
Packaged with this daemon are:-
Dan Bernstein's authuser library
Dan Bernstien's test program for the original authd
Code using identd and authuser can be found on 'ftp.lysator.liu.se'
in the directory 'lib/ident'. You will also find a version of Peter's
pidentd which will work with other architectures. You can also look
for Dan's authd package (available in several flavours) which may
include support for your machine!
Nigel Metheringham <nigelm@ohm.york.ac.uk>
Department of Electronics
University of York
England.
Why was this release made? Well, after the code had been out there for several months, Carl Edman released this neat new version of finger, that supports RFC931 among other things. Suddenly there was a flurry of people getting that, and as a side effect installing nidentd servers on their machines. This was when I got my first comments about the software, and guess what, I had fluffed in the install section! I run a network of NeXTs as part of a larger homogenous network of systems. One of the first things I do as part of my installation procedure is niload the proper (ie matching our other systems) group file up into the local netinfo db - otherwise our mail and news gets messed up. However when I produced this package I forgot this, and assumed that my group names are correct. What I think of as "sys", you all call "kmem", and this totally wrecks my install instructions. Obviously many people worked this out, but this extra release is purely to correct the bugs in the install instructions - nothing else has been touched (other than a setgid option in the Makefile). So whats going to happen. When I get time (some hope), I will update nidentd to include all the neat additional features that Peter Eriksson has put in his version - basically it will run permanently as a daemon rather than being kicked off by inetd, and it will hold the kernel information internally which will mean it will respond *much* faster. I am also watching the development of the TAP protocols (basically RFC931 taken onward), and will keep it up with that. I will try to keep it as small and simple as possible within these constraints! Nigel Metheringham <nigelm@ohm.york.ac.uk> 24 July, 1992
These are the contents of the former NiCE NeXT User Group NeXTSTEP/OpenStep software archive, currently hosted by Netfuture.ch.