/usr/share/doc/nmap-ncat-6.40/ Current File: //usr/share/doc/nmap-ncat-6.40/README
   .       .         
   \`-"'"-'/
    } 6 6 {    
   ==. Y ,==   
     /^^^\  .
    /     \  )           
   (  )-(  )/     _
   -""---""---   /
  /   Ncat    \_/
 (     ____
  \_.=|____E


README for Ncat 
---------------

Ncat is a reimplementation of the currently splintered and reasonably
unmaintained Netcat family.  Ncat will do pretty much everything that
all the other Netcat's do, all in one place.  Plus it has the added
benefit of spanky new features and ongoing development.

Ncat was designed with the original Netcat interface in mind.  Rather
than replacing the old Netcat interface with a brand new (and thus more
convoluted) set of options, the Ncat interface was intentionally kept
clean and simple to use, just like the original product.

Ncat provides most of the features present in the original Netcat but with
a complete overhaul and rewrite, along with completely new features and a
combination of other well received features of other Netcat products, such
as IPv6 and SSL support.

The port scanning support has been entirely removed from Ncat.  The
reason for this is fairly obvious: there is a better port scanner out
there already... :)

Ncat can act as either a client or server, using TCP or UDP over IPv4 or
IPv6.  SSL support is provided for both the client and server mode.

There is also a new "connection brokering" feature which enables two or
more hosts to connect that previously were unable to directly communicate
with each other.

For example: "Host A" can connect to "Host B" but not "Host C"
             "Host C" can connect to "Host B" but not "Host A"

It is clear, then, that if you could connect to "Host B" then "Host A"
and "Host C" could directly communicate...

[HostA] <------> [HostB-with-Ncat-Broker] <------> [HostC]

Ncat's connection brokering will allow you to connect between Host A and
Host C via Host B without the trouble of having to have SOCKS support,
etc.  This is still somewhat experimental behaviour.

Ncat has support for HTTP "CONNECT" via an HTTP proxy server such as Squid.
It can also connect via a SOCKS4 server and is very flexible in terms of
how it shuffles your data around.

Ncat can also spawn its own HTTP CONNECT proxy server for your
own relaying requirements.

Ncat has the ability to execute a program and handle the I/O for its data
over the socket.  In other words, Ncat can "add" networking support to
applications that currently have none.

For example, you could:

ncat --exec "/bin/bash" -l 5000

NOTE: This is exceptionally dangerous behaviour, as it leaves an open shell
sitting directly accessible to anyone who is able to connect to port 5000.
See the allow and deny options below for securing your Ncat processes.

Ncat has a TCP and UDP "redir"-style redirection feature to allow the user
to redirect traffic from one host to another.

For example:

ncat --exec "/usr/local/bin/ncat www.example.com 80" -l 8888

This command binds Ncat to the local machine on port 8888 and redirects
connections to www.example.com.  You may also find uses for this as a "host
hiding" system.  Similar to SOCKS4 but without any of the hassle of having
to have SOCKS support in the application.

This also begs the question of, "What would happen if you decided you wanted
to pass the --udp flag in to the above command somewhere?"

In this case, you would have a TCP to UDP "gender changer". 

If you have an application that only makes only TCP connections, for example,
you could spawn a Ncat process to listen on a the applications TCP port and
then redirect the TCP connection out to the final destination, only over UDP.

The --allow and --deny options are provided to prevent unauthorized access to
any Ncat process that is listening on a port.  These options are also paired
with the --allowfile and --denyfile options, similar in behaviour to Nmap.

The allow and deny options accept a number of different IP address formats
for maximum flexibility:

A single IP address, of the format:

	ip.ip.ip.ip

EG: 192.168.10.1

A CIDR-style IP address range, of the format:
 
	ip.ip.ip.ip/cidr 

EG: 192.168.10.0/24

An IP and full netmask, of the format:

	ip.ip.ip.ip:nm.nm.nm.nm

EG: 192.168.10.4:255.255.255.255

An IP address with wildcards:

	ip.ip.ip.*
	ip.ip.*.*
	ip.*.*.*

These rules may also be used in a flat file, delimited by newlines.  An
example of a full ACL is included in docs/examples/ labelled 'iplist'.  Also
note that comments start with a # and are perfectly acceptable for use in
the ACL's.

For example, the file "ipaccess" might look like:

# Abuse from ADSL user.
88.223.14.1/32

# This guy is scanning us for SOCKS4 servers to abuse, block his /24
194.213.167.*

To implement this IP address ACL simply run:

ncat --denyfile /path/to/file/ipaccess -l 7000

For further documentation, please see the man page.

--Chris Gibson and Kris Katterjohn