----- Original Message -----
From: "Lee Elliott" <[EMAIL PROTECTED]>
To: "FlightGear developers discussions" <[EMAIL PROTECTED]>
Sent: Saturday, July 31, 2004 8:46 AM
Subject: Re: [Flightgear-devel] Data source battle?
> On Friday 30 July 2004 23:19, John Wojnaroski wrote:
> > When using real-time weather via the net and the native-ctrls interface
> > input control data to FG it appears both "sources" attempt to write to
> > environmental properties; this is most noticeable with the wind vector
> > discontinuities in the ground speed
> > "turning off" the specific wind properties update in native-ctrls clears
> > the conflict but restricts wind updates to the real-time system.
> > Has anyone experienced this problem?
> > Regards
> > John W.
> This sounds like it may be at the bottom of some problems I've been having
> with weather.
The problem occurs when using the native-ctrls sockets and protocols to send
a control packet from the sim hardware to FG at a 24Hz rate. The
native-ctrls.cxx updates the property tree with the values in the packet
which may or may not be set by the sender. On the next frame, the real-time
weather kicks in until the next data packet arrives. If you need or want
only part of the packet you have to hack the code. There is a section of
code in native-ctrls.cxx around line #386 that updates the property tree
with the values in the received packet. Normally, this function should not
run unless you've enabled the native-ctrls protocol and socket.
Curtis posted an email about two months ago soliciting some ideas and
thoughts on the whole topic of the interface between the FDMs, FG, and the
network. Rework in that area would be a major effort. Especially if the idea
is to provide support for a network of machines running portions of a
simulator such as aircraft subsystems, displays, hardware driver interfaces,
FDMs, visual systems, etc
Flightgear-devel mailing list