Geoff McLane writes: > Yeah! Precisely. perfectly. Can I give you the number of my > person-in-charge-of-upgrade of our in-house major system? ... and you > tell them (actually her) - err, I only write an fdm for that in which in > can > run. Get a better capability, or ??? shut up ??? :-)))
There have been several concrete, substantive posts describing the minimal flightgear machine. Why don't you forward those along to her and if she has any further questions, we can try to address them as well. > > > only 2 or 4 MB of video RAM I am surprised you can run at all. > > It does run. It runs well. It is the best since sliced bread! Keep it > up forever. I am really for dissemination of knowledge. And being able > to 'magic carpet' along a coast line, could really help in geography > lessons ... at the very least ... > > There are really and truely not many (free or otherwise) products out > there > where you can elevate your self say just 5000 feet above mean sea level > above SF, turn south, take on some speed and travel along the 'real' usa > coast until say you reach the LA basin, which you 'recognise' from your > geography lessons, or on further south ... forever, until the tiles run > out !!! > > So why does this FGFS workable-in-this-tiny-pc-environment scenerio > become 'impossible', or at least more/very difficult, if you add in the > EOM, 'equation of motion' of an aircraft? You tell me? Have you compared JSBSim against LaRCsim and YASim? > I have just agreed i turn OFF another big graphic grabber - the panel - > and run only on the HUD - modified as earlier mentioned to ALWAYS > show frame rate ... HUD's must do this ... maybe a hud xml changes? > > The only 'suggestion' i had was that an fdm who did well in the overall > scene would have *nothing* to do until eng/mags/wind became 'alive. No > trim or ANYTHING until at least all the scenery tiles had been laid ... As > mentioned, that can be several 'minutes' into the ride ... The FDM decides the motion of the aircraft. There are always forces acting on it, unless you figure out a way to turn off gravity. I just don't think this optimization would be that useful. Minutes to finish loading the initial tiles? Are you loading from a really slow cd-rom or a slow network share? On my machine with a local hard drive, the tiles are loaded within a second of when the graphics come up. > I like JSBSim, at least in its c172 'feel'. I am not so familiar with > other ac, > but spend at least some of my time c172'ing around airfields. But this > ac is not a real mover ... and must try others ... > > I am sorry none of what i said made any difference ... but it is ok. i > enjoyed saying it ... > > A few questions ... some 'silly' cvs things ... > > I use the cmd cvs update -dP, but in FlightGear, i constantly get some > 'warning' report about a missing 'yasim\util'??? directory ... but is says > 'skipped' ... Is this all ok? I thought -P = prune empty. But of course > it will NOT be empty until you/i/cvs delete things ... > > Now i am also getting the same thing in SimGear regarding the dearly > recently departed zlib directory ... You mean cvs DOES NOT delete > non-existant items in what is now knows as a non-existant directory? > > Then that seems to miss the point of 'update' ... CVS does not provide for deleting directories from the repository, but I break the rules and do it anyway. Go into the CVS/Entries and simply delete the item that is being complained about. > I certainly agree 'certain' iterative memory allocation should have > 'no impact' on a well written memory system ... did i not mentions i > run win32? :-)) I don't think the problems are necessarily win32 ... I prefer to run Linux of course ... but there are many people out there running win32 without much trouble. Are you configuring --with-logging or --without-logging? Any console output can *REALLY* kill performance of a win32 app, perhaps that's what you are seeing? Regards, Curt. -- Curtis Olson IVLab / HumanFIRST Program FlightGear Project Twin Cities [EMAIL PROTECTED] [EMAIL PROTECTED] Minnesota http://www.menet.umn.edu/~curt http://www.flightgear.org _______________________________________________ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
