Hi, Ritesh Sood escreveu: > Hi all, > > This mail is more of a feature request, and looking at the number of > messages on the web, I'm sure quite a number of users would be happy to > have this functionality, which is already provided by many commercial > Xservers for windows. > > I want use Xephyr/Xnest on my home machine local_host (as display :1) > and have the display controlled by xdm running on a remote application server > (app_server) > > First, please have your browser's font set to a monospaced one so that > the boxes below are displayed correctly. > > Here's how the network "topology" looks like. > > +---------------+ +------------+ +-------------+ > | local_host | | NAT server | | app_server | > | 192.168.0.100 |-------> | 1.2.3.4 |--------->| 5.6.7.8 | > | running Xnest | | | | my.univ.edu | > | my.univ.edu | > | on display :1 | | my.isp.com | | running xdm | > +---------------+ +------------+ +-------------+ > > At the app_server end, Xaccess contains > *.univ.edu NOBROADCAST > *.isp.com NOBROADCAST > to have some measure of security > > I'm running xdm as > # xdm -debug 1 -config .... > > Within the university network of-course, things work very well. From > local_host too, at-least XDMCP authentication is happening correctly, > i.e. xdm sees that the incoming request is from *.isp.com. and considers > it legitimate. > > Next, it tries to open 192.168.0.100:1 for login window, etc; and that > of-course fails. > > Just to make sure that port forwarding on 60xx ports is happening correctly, > I do > $ xterm -display my.isp.com:1.0 > and that works alright. > > As i mentioned above, many Xserver implementations for windows provide > an option so that the NAT IP address can be passed to xdm instead of > XDMCP picking up the local_host address by default. See these FAQs, for > instance: > http://connectivity.hummingbird.com/support/nc/exceed/exc9003009.html?cks=y > http://www.netsarang.com/products/xmg_faq.html > > It would be great if we could have similar functionality in the Xorg > Xservers.
Yeah, I would like this kind of feature some time ago as well but seems that our world is finally (not so quickly though) turning to IPv6 [0]. There would be another crazy idea to traverse NAT using hole punching technique. Follow this link: http://vignatti.wordpress.com/2008/03/21/traversing-x11-clients-behind-nat-or-x11-end-to-end-connectivity/ [0] people found another motivation besides the lack of address space which is the energy saving. Seems that NAT must send a "keep alive" message every 30-180 seconds to keep the address and connection active. It can consume a significant amount of energy, specially for mobile devices. Cheers, -- Tiago Vignatti C3SL - Centro de Computação Científica e Software Livre www.c3sl.ufpr.br _______________________________________________ xorg mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/xorg
