Durk Talsma > Sent: 07 July 2007 08:22 > To: FlightGear developers discussions > Subject: Re: [Flightgear-devel] Today's CVS > > > Hi Vivian, > > On Friday 06 July 2007 23:31, Vivian Meazza wrote: > > > > After further testing I can confirm that the disappearance > is readily > > repeatable. If you position your ac on the threshold of 28R > at KSFO, > > as the AI aircraft are about to trample all over you they > obligingly > > commit suicide. The first then reappears well down runway > 28L taking > > off, the second just disappears. > > This I can replicate! I was unable to reproduce this, until > you provided a > vital clue in the current message: "as they are about to > trample all over > you". I usually try to get off the runway ASAP using the UFO, > when inspecting > traffic behavior, so I missed this one completely. Initially, > I was a little > confused about the "teleporting" part of your bug report, but > it makes sense > now. You couldn't know, but the disappearing and "respawning" are two > independent phenomena. > > There is some basic proximity detection code which cause AI > aircraft to wait > for the user's aircraft. More recently I added a new function > that detects > situations in the ground network were aircraft are eventually > waiting for > themselves. (i.e. a waits for b; b waits for c; but c waits > for a). The > occurrence of these is not frequent, but when it happens, the waiting > aircraft can block all other traffic. Eventually, this needs > to be resolved > more gracefully, probably by moving one of the offending > aircraft back, or by > implementing more sophisticated "hold position" instructions. > Until that time > the offending aircraft are removed from the scene. > > It seems this function is triggered whenever an AI aircraft > is waiting for the > user. I believe a solution is fairly straightforward, but > requires some more > testing. I'll probably have something at the end of the day. > > The "respawning" on the other hand, is a different phenomenon. During > initialization, the traffic manager makes an estimate as to > which stage of a > flight an aircraft is: Wait at gate / push back / taxi / take > off / climb / > cruise. When an aircraft is considered to be in the take off > stage, it is > placed on the runway, and given take off speed during > initialization, thus > creating the illusion of a spontaneous spawning of these > aircraft. It's not > really an ideal situation, waiting for refinement.
OK so it's 2 features, not 1 bug - excellent. I also noted that the route taken from parking to active runway seemed a bit odd, but then I compared our taxiways to those on Google Earth - we seem to have bits missing alongside 28L, which explains that. Not good for our default airport though. > FWIW, I also hope to have a look at the tanker speeds this weekend. Your proposed fix seems to work here, on the basis of just one test, I haven't had time to do the job properly. Vivian ------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/ _______________________________________________ Flightgear-devel mailing list Flightgear-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/flightgear-devel
