David Megginson writes:
> Curtis L. Olson writes:
> 
>  > I'm flying the stock JSBSim C310 ... Linux debian, nvidia card,
>  > gcc-2.95.4, compiled with threaded scenery paging enabled, yadda,
>  > yadda ...  Am I just suffering from heat stroke now that we had an 80
>  > degree day after a long MN winter?  Has anyone else seen this?  I'm I
>  > the only one that uses the nav radios? :-) Anyone have any ideas or
>  > suggestions?  I think this problem has been lurking for quite some
>  > time.
> 
> I'm not using the nav radios much these days for two reasons:
> 
> 1. They're hard to use from inside the 3D cockpit, at least until Andy
>    gets mouse clicks working on his wrap-around panel; and
> 
> 2. All of my training right now is VFR, so that's what I'm
>    practicing.
> 
> It's hard to imagine what might be causing the problems.  Norm sent in
> some panel speedups a while ago, but I'm not sure I committed them,
> and I don't think they touched the text stuff anyway.  The text
> instruments use sprintf, so maybe something has screwed up the library
> somehow.

David,

I looked a bit at the text drawing code in panel.cxx and I'm almost
certain I see where the problem is happening, although I need to make
it fail to get more information on why/how.

The text drawing routines draw the text every frame, but only
update/change the contents the data when _now - _then > 100000
(i.e. 10x a second.)

I strongly suspect that this time difference calculation is getting
out of whack somewhere somehow.  Maybe one of these times is getting
corrupted some how so _then is in the future and the answer is always
< 0?  Maybe we are overflowing a value in our calculations and getting
the wrong answer in some situations?  Maybe this happens if we stay in
a nonpanel view for too long?  I don't know why, but I very strongly
suspect this is where the problem is.

I'll run for a while and see if I can figure out how to make this
break, and dump out the actual time values and see if I can figure out
what's going on with them.

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

Reply via email to