I've got a few of John's and some of World AI's plans installed and Johns plans make sense. Overnights are accurate- right now using World AI, I have Boston fully populated at 7am (I mean almost all gates are filled with DAL or USAirways jets) none of the other carriers that should be there! I also have Boston roughly configured for the correct number of gates, but there are some space constratints Especially in Terminal B- no new Usairways gates, and the AA side is not big enough.Tim

Tim,I really did download 'real life' data from the DOT. The overnights should be accurate :) ... with two 'minor' exception.1) at the end of each week it was required to move the aircraft back to its originating airport. most often this is done with a flight number of 0.2) from time to time it was apparent that the airline 'deadheaded' an aircraft (flew it empty) from one location to another as it is listed with DOT as having flown (for example) a flight from KMSP to KMEM at 8:00 then the same aircraft would fly from KMSP to KDTW at 3:00, with no flight listed from KMEM to KMSP. These deadhead flights as well (for the most part) were given a flight number of 0.-john

That's the beauty :) Using the Dot data and a weekly sample you can follow the aircraft around on a week's flying with the appropriate linearity that would be in place with a finite fleet of aircraft. With the World AI, there are point-to-point flights that have an inordinate amount of time at the gate. This is OK for some schedules- Virgin's long haul routes come to mind, but these palnes are used for in excess of 10 hours a day on average.IN the case of US Airways, there are few enough international destinations that filling in a complete schedule would be fairly easy :)Tim