Re: [Flightgear-devel] Possible cause of JSBSim airport bug

[<js...@ha...>]

> It would be a good idea to make JSBSim robust enough to deal with this
> kind of problem, but for now, we need to fix current_elev_ssg,

I completely agree. Off the top of my head, however, I am unsure how to address 
this - which is partly due to my ignorance on how the scenery issues play into 
the initialization issues and how/when it works with FDM init. One thing to do 
would be to sense when the aircraft contacts the ground in an obviously 
incorrect way or at any speed that results in obviously unsurvivable 
velocities, and mark this is a crash, and set values accordingly or freeze or 
whatever. If it turns out that JSBSim is correctly placing the aircraft at 
altitude with no velocity and the aircraft merely falls out of the sky, we 
(Tony and/or I) need to make sure the resulting "farm buying" doesn't result in 
seg faults or NaNs. I wonder if greater use of try/catch blocks would be 
beneficial?

> preferably for 7.7; when that's done, on-ground starts with JSBSim
> might start working again.

This would be welcome. Debugging this one couldn't have been fun. I much 
appreciate it, and also welcome ideas/suggestions/flames on what we need to do 
to make JSBSim more robust in this respect.

Jon