Lee Elliott wrote

> Sent: Tuesday, September 21, 2004 8:22 PM
> To: FlightGear developers discussions
> Subject: Re: [Flightgear-devel] Problem with ballistic sub-model
> On Monday 20 September 2004 08:39, Vivian Meazza wrote: {snip...]
> > Lee,
> >
> > Here's some pics off the effects I obtained:
> >
> > http://myweb.tiscali.co.uk/vmeazza/FlightGear/zero_drag-full_g
> >ravity.jpg
> >
> > http://myweb.tiscali.co.uk/vmeazza/FlightGear/zero_gravity-acc
> >elerating.jpg
> >
> > http://myweb.tiscali.co.uk/vmeazza/FlightGear/zero_gravity-bom
> >bs_overtaking. jpg
> >
> > http://myweb.tiscali.co.uk/vmeazza/FlightGear/zero_gravity-bra
> >king.jpg
> >
> > I hope that you can achieve the same.
> >
> > Regards
> >
> > Vivian
> Sadly, no.  Those pics are pretty much what I was hoping to see 
> here but it's just not happening.  This is very odd.

I'm very encouraged that I got what you expected.
> I could post a pic showing the a/c after braking, with the string 
> of bombs behind and the cluster around the nose (which would 
> show that it had moved, while releasing bombs and was now 
> stationary) but is there any point?

It might help in diagnosing the problem, and any others you think might
help. Include your settings of eda and cd, please
> I'm a bit reluctant to do a full fresh cvs checkout because I'd 
> have to also get a new base package too, wouldn't I?  Still on 
> dial-up here so I'd like to avoid having to do that if possible.

I felt that guestimating the value for eda was too difficult for widely
differing object such as bullets, drop tanks and bombs (big ones) so I have
asked Erik to commit a change to cvs that allows direct entry in the
submodels.xml file of the raw data for cd, weight and cross-sectional area.
Works for your bomb here, and should enable you to eliminate one source of

> If I delete the source code directories from my local cvs and 
> then did an update, would it replace anything that was missing?

Yes, that would work (and I do it in desperation when I've lost the bubble
on changes)  but it you don't gain anything - cvs only downloads what is

> I guess I could try getting FG installed on one of my other 
> workstations and see if I get the same results.

Good plan to diagnose the problem
> I did try setting a more verbose log-level but didn't see 
> anything there either.

> This is such a strange problem.  Because there are no apparent 
> run-time errors it's difficult not to conclude that there's a 
> logic error in the code but the fact it works elsewhere rules 
> that out.

FWIW - you seem to be the only one suffering here. David Culp has also
produced some nice pics of you're your bomb behaving more or les correctly.
I say more or less because I feel we need to handle air density more
accurately: at the moment we use the sea-level value for all altitudes. Good
enough for demonstration purposes only. 

> Feeling pretty stumped here:-/

If you really do have the latest code, then the problem may lie in the data,
in particular in the submodel.xml file. The math is very sensitive to
changes in cd and eda. You can get the submodels going in some pretty
unexpected direction quite easily, and it's sometimes quite hard to find
just where they are going.



Flightgear-devel mailing list

Reply via email to