11/8    is this supposed to work: open a WinXP remote desktop connection from A
        to B, and from B remotely open one back to A?
        \_ Probably not.  Why would you do this?
        \_ You want it to recursively draw the screen?
        \_ You can go A -> B- > C -> A.  Not sure why....
           \_ How does the screen look like???  Does it recurse infinitely?
              \_ It *very* quickly turns into a blur.  There's only so many
                 pixels.  Since the system has to reduce pixel count on each
                 iteration/display you very quickly get to the point where
                 it only uses a pixel to represent an entire screen.  I guess
                 you could go into full screen mode, get rid of the task bar
                 and other stuff that takes up a static amount of space but
                 I've never bothered.  Go get 3 machines and let us know.
                 \_ When you are sitting in front of A and have connected
                    A->B->C, and you try to connect A->B->C->A, shouldn't it
                    lock the whole screen as soon as you've entered the
                    password for A since A is now being accessed remotely from
                    C?
                    \_ But... but C is remotely controlled from A?? The
                       universe will explode! Actually I think I'll try this
                       experiment with my cubicle neighbors.
        \_ Dunno about remote desktop but VNC certainly will let you link a
          client to a machine listening to the client.  I did it once.  was a
          somewhat amusing ride. -ERicM
        \_ FYI: the reason I asked is because from home i wanted to make sure I
           could remotely connect from work through my firewall etc. So I went
           through the VPN and remote'd to work, from there remote'd to home,
           and it would get to the login screen and then my VPN died, every
           time. Something wasn't happy.
        \_ The remote desktop server in WinXP is a trimmed down of full-blown
           terminal services.  The main difference is that the WinXP RDS only
           supports one connection at a time.  If you connect to a computer,
           it locks the console and takes control of the desktop session.  If
           the system was already under the control of a remote desktop
           session, the new session will take over the old one.  Therefore it
           is impossible to create an infinite loop of desktops within WinXP.
           I'm pretty sure that the same goes for full terminal services
           (except that you can't kick off the console user, instead it creates
           a new desktop under terminal services control).