Hi, On Friday 06 January 2006 00:37, Lee Elliott wrote: > On Thursday 05 Jan 2006 22:46, Curtis L. Olson wrote: > > Also, who ever is developing and working on the MP and AI > > code: > > > > PLEASE, PLEASE, PLEASE, PLEASE, PLEASE, PLEASE, PLEASE, > > PLEASE, PLEASE, PLEASE, PLEASE, PLEASE, PLEASE, PLEASE, > > PLEASE, PLEASE, PLEASE, PLEASE, PLEASE, PLEASE, PLEASE, > > PLEASE, PLEASE, PLEASE, PLEASE, PLEASE, PLEASE! > > > > Please consider users with multiple display channels driven > > from multiple computeres synced together over the network. > > > > Nothing is more embarassing (well maybe I shouldn't say > > nothing) than demoing the new sim software on a 6 screen > > visual system flightgear based simulator and having 6 > > independent ATC/Tower communications sessions going (one on > > each screen) and then having different airplanes appear on one > > screen and not on any of the others. > > > > Please consider some way to replicate the MP/AI traffic across > > multiple visual channels, perhaps slave from one master > > machine that communicates with the outside MP server. > > I was just wondering if pigeon's FG map web-site could be adapted > to function as a radar and had started to think about how it > might be nice to include AI traffic as well when the problem of > inconsistent AI aircraft that you mentioned above occurred to > me:)
The problem with inconsistent AI objects should not exist. All we have to do is: make sure that there is only one source for every AI object. That means, that one instance of flightgear might feed 10 AI objects to the virtual world, but this instance is authorative for those objects. There are no other feeders for them. Although we might invent some failover mechanism, I think we shoot ourself in the head if we try to implement multiple feeders for one object and try to keep them in sync. > Perhaps a possible solution would be for the MP server to > generate AI traffic and for local AI traffic to be switched off > when connected to an MP server. In fact we could implement simple feeders which do nothing more then feeding some objects to the virtual world. But I would prefer to have a dedicated instance of flightgear to be the feeder, as it will reduce code replication. For ground objects (cars, trains whatever) the feeder will need terrain information, so why not use flightgear itself (without a visible user interface like openGL)? > For the multi-system type set-up you're using for your 6 screen > configuration there would have to be some way of nominating one > of the systems to take on the MP AI role, with the remaining > systems treating it as an MP server. From my (multiplayer) point of view, there is no change whatsoever regarding multi display configurations. One instance of flightgear does the communication with the virtual world (server) and feeds only views to the remaining systems just as it is now. regards, Oliver ------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_idv37&alloc_id865&op=click _______________________________________________ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
