cvs login fails (connection refused) - why?

[joeshmoeisnotavailable]

I am trying to setup a CVS server on a Fedora Core 8 Linux box.

In order to enable the cvs pserver, I basically modified /etc/xinetd.d/
cvs to contain:

service cvspserver
{
        disable         = no
        port            = 2401
        socket_type     = stream
        protocol        = tcp
        wait            = no
        user            = cvsroot
        passenv         =
        server          = /usr/bin/cvs
        env             = HOME=/home/cvsroot
        server_args     = --allow-root=/home/cvsroot -f pserver
        log             = /var/log/cvslog
        id              = cvspserver
#       bind            = 127.0.0.1
}

I also configured the firewall to allow both tcp and udp on port 2401
(cvspserver).

I also created in $CVSROOT/CVSROOT the password file, passwd.

On the Windows (XP) client, I added the environment variables;

LOGNAME = joeshmoe
CVSROOT = :pserver:address@hidden:/home/cvsroot

I then try to do cvs login from the client:

~> cvs login
Logging in to :pserver:address@hidden:2401/home/cvsroot
CVS password:

But when I type the password (which I triple-verified!), cvs login
simply fails:

cvs.exe [login aborted]: connect to 192.168.0.2(192.168.0.2):2401
failed: Connection timed out

My questions is: How do I troubleshoot such a problem? How do I know
why it failed and how to fix it?

BTW, I tried telnet 192.168.0.2 2401 but all I receive in response is:

Connecting To 192.168.157.2...Could not open connection to the host,
on port 2401: Connect failed

What does this mean? Is it possible that something totally unrelated
to cvs blocks this connection? SELinux perhaps? How do I find out?

Thanks,
Joe