|
From: John W. <ca...@mm...> - 2001-09-06 18:59:54
|
Alex writes: > Have you checked whether the OS is actually doing the broadcasts using another > program to send and receive the packets ? I've had some Windows machines that > have ignored my specification of the broadcast address and used their own. > As a result, the address I thought I had to use for broadcast ... didn't. > That's a good point. I've got windows95 (2), ME (1) , 2000 (1) and linux (1) on four machines I'll try a variety of mixes for servers and clients and see what happens. The network functions are built into the linux kernel. I have no idea what Gates and MS have done! Not sure I want to find out ;-) > In the short term, try just repeating the native-out parameter and use > point-to-point UDP links for every client until you have everything else > working. Less efficient for the server, but probably not enough to care. Not sure I follow ... would you restart fgfs each time with a new "out" address for each client (in this case three). The socket 5500 binds to the IP address when the socket in instantiated, you would have to create additional socket/port bindings inside fgfs to do this simultaneously. I have had the 3 "server" (sometimes 4 when I can "steal" the spouse's) machines receiving datagrams so network connectivity appeares okay. I have not tried to run any of these machines as the client. that might be worth a look. Based on the results of running some tests with Norm's socktest and some utilities I have the data send by the client (the one running the fdm) is getting on the LAN with the correct IP address, whether unique or broadcast. I'll try your suggestions and then follow the "cookie crumb trail" and see if I can find where and how the bits get from point A to B. BTW: An interesting aspect of this feature (I'm sure some one has thought of it) is that if you could drive the 3d aircraft models with an external source then two or more players could then fly "formation" with each other. Regards Jack W > > > _______________________________________________ > Flightgear-devel mailing list > Fli...@li... > https://lists.sourceforge.net/lists/listinfo/flightgear-devel > |