On 2/1/07, Joacim Persson wrote:
Affirmative. ;) Looking at the code I can tell there's years of
development
in layers on top of each other there, not entirely consistent with a one
and same philosophy.
Actually the code was mostly developed in one chunk, it just came out extra
complicated ... not sure why that was. I always figured that my problems
getting my head around it were mainly due to me not seeing the code from
exactly the same perspective as the original author. Still, it seems like
it could be cleaned up substantially.
Time interpolating over several minutes on a per plane basis will make
pilots on MP each flying in their own weather, even when they are all
using
METAR weather (which IMO should be mandatory on MP). So they wouldn't even
agree on where the flighlevels are.
But unless you somehow enforce that all active pilots update their weather
in lockstep (which we don't do know, and is probably a bad idea from a
network usage and net friendliness perspective) you will always have this
problem. We have this problem now.
I think the solution to that would be to add weather conditions to the MP
data exchange format so the MP server tells you what your weather is and we
can control it at that level.
That would mess upp the IVAO stuff Pep Ribal was talking about here
recently.
But now you are talking about flying with MSFS and X-Plane users. I must
assume that both of them do their weather updates differently. Mismatches
in weather and terrain are always going to be an issue when you put
different vendor's simulators together.
As I mentioned in my last message,
I'm more into using a combination of geometrical interpolation and time
interpolation using a small time delta (seconds) to filter out the
remaining transients. But again, we can't get enough data from METAR only.
(Think of intercontinental flight)
I can also imagine that different users with different needs for a flight
simulator would have different needs for weather model. Someone might want
to
have a detailed simulation of microweather using software like MetPhoMod
(http://www.giub.unibe.ch/klimet/metphomod/, GPL'ed), for simulating
flight
in a mountain area with difficult up- and downdrafts; soaring pilots want
realistic thermals that a motor pilot only thinks of as "turbulence".
Helicopter landings on offshore oil rigs etc etc. My feeling is that there
should first and foremost be a defined interface for weather vs the rest
of
the sim system. Then people can hook up whatever they like to it using
that
interface.
Well for the most part, we have an interface via the property system. We
should be able to put any back end behind it to update those properties
anyway we want ... but this is a distraction from your original point of the
problems of not having everyone in a MP system update their weather in lock
step and in exactly the same way.
I am currently time interpolating everything weatherish typed as a double
using a modified fgSetDouble() with an extra optional argument. (But for
some
bizarre reason, the sea-level air pressure -- what I particulary want to
smooth out -- refuses to be interpolated yet. AAaargh!)
Yeah, if I had time to look at the weather stuff right now, I would have
done so, the discontinuous updates are not attractive. I appreciate you
taking the time to dig into this. I'd encourage you to start simple (which
you appear to be doing) and add complexity later ...
Horizontal cloud density is not a double though -- conceptually it's an
enumerate rather. Not much I can do about that. But the clouds doesn't
knock the aircraft over in a tunnel roll.
When a weather update only changes the cloud layer heights, then we can
slowly/smoothly move them to the new heights. But when we need to add or
remove layers or the coverage % changes, then we might be able to slowly
blend layers in and out using alpha/transparency.
Regards,
Curt.
--
Curtis Olson - University of Minnesota - FlightGear Project
http://baron.flightgear.org/~curt/
<http://baron.flightgear.org/%7Ecurt/> http://www.humanfirst.umn.edu/
http://www.flightgear.org
Unique text: 2f585eeea02e2c79d7b1d8c4963bae2d
-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier.
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
_______________________________________________
Flightgear-devel mailing list
Flightgear-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/flightgear-devel
