I'm not sure what you're asking--I think you're saying you want to be able
to host an applet such that when a user uses it, it can show them (and the
server it came from) their public IP address even if the client is behind an
anonymous proxy?

An anonymous proxy rewrites incoming requests and resubmits them to the
actual server elsewhere. One request will be to retrieve the applet and
another if the applet makes some kind of request on its own. If the applet
(running in the client's browser) is making requests to its code base
server, the code base will be the proxy server who will be able to see the
client's public IP (but not care) and the requests will simply be re-made
from the proxy which will instead appear as the request source.  If the
applet makes the requests directly to its home server (built-in server
name), the requests would (if they got through) would show the applet's
public IP address.  However, these requests will cause a security exception
because the applet can only talk to the server from which the browser loaded
it.

So the answer is, I don't think it's possible with an applet unless you get
the user to give permission (via signing, policy, etc.)  Note that there may
be bugs that make this possible, but that's another matter.

Cheers,
Matt Humphrey  matth@ivizNOSPAM.com  http://www.iviz.com/

Wondering what your actually looking for, there are 3 possible IP
addresses, depending on the setup of the access machine.

1) The proxy server
2) The firewall
3) The actual client machine

If you have a proxy server, you probably also have a firewall, without a
firewall there is no need to use a proxy server.  Firewalls (and
routers) often use a technique called NAT (Network Address Translation),
which means that you only need 1 public IP address for potentially
thousands of machines.

For example I have 3 machines here, Leafs, Habs, Sens, they are
connected to a 4th machine the firewall.  The firewall is an old PC
running Smoothwall Express 2.0, which has all of the updates applied
(except the last one, which I downloaded this morning).

Now my setup uses DHCP because I find that using that is easier then
trying to manage IP addresses, even though there are only 4 machines
that can be physically hooked up.  However the pool of addresses it uses
are between 192.168.0.2 and 192.168.0.200, addresses which are reserved
for local network only use.  The firewall uses NAT to make it all work.

Now back to the original question:

Any applet that gets the Firewall address probably can't do anything
with it, any applet from outside the firewall that gets the internal
address is looking at junk.  Proxy addresses probably don't mean much
either.  If your inside the firewall, just ping the machine name.

W