New Weather Interface and Observations

[pilotjohn]

Pete,

Looking for some additional info on NWI.

1. What is the tops setting, it doesn't seem to make difference?

2. Why are there 10 cloud types, but only 1, 8, 9, 10 seem to be used?

For others, I empirically determined some cloud/precip. rendering rules (maybe it'll be helpful to some):

Cirrus: Thin no matter what.

Stratus: Half as thick as requested.

Cumulus: Less than 2000 puffy, otherwise dark base, more than 15000 towering.

Thunder: Less than 15000 towering, otherwise as requested.

P-base: Write 10x, will not be lower than real top of layer below.

[Pete Dowson]

1. What is the tops setting, it doesn't seem to make difference?

It does in earlier versions of FS -- in FSX there's no way to specify cloud thickness. Instead it is worked out by some algorithm internally according to cloud type and base altitude. Very annoying, as you can't read the cloud thickness either, so you cannot really determine programmatically whether you are in cloud or not!

2. Why are there 10 cloud types, but only 1, 8, 9, 10 seem to be used?

That's something you would have needed to ask the Microsoft FS team. I assume that when they originally set the type numbers they were going to implement a whole lot more.

For others, I empirically determined some cloud/precip. rendering rules (maybe it'll be helpful to some):

Cirrus: Thin no matter what.

Stratus: Half as thick as requested.

Cumulus: Less than 2000 puffy, otherwise dark base, more than 15000 towering.

Thunder: Less than 15000 towering, otherwise as requested.

P-base: Write 10x, will not be lower than real top of layer below.

Thanks. But how are you measuring the results? If you are using the values returned by FSUIPC, they are arbitrary guesses. The only way to measure if slew up through the layers and check the alititudes.

Pete

[pilotjohn]

Thanks. But how are you measuring the results? If you are using the values returned by FSUIPC, they are arbitrary guesses. The only way to measure if slew up through the layers and check the alititudes.

It was determined empirically, slewing up/down using many combinations of settings.

A few more questions... As I mentioned, I'm working on a Lua Random Weather generator.

I'm using vstruct to pack/unpack structures and need confirmation that chars are 1 byte, shorts are 2 bytes, and ints are 4 bytes, as FSX/FSUIPC is compiled. Is this correct?

Also, NewWeather.h in the SDK is different than NewWeather.h in New Weather Interface for FS2004.zip. Which one is the correct one for FSX?

[Pete Dowson]

I'm using vstruct to pack/unpack structures and need confirmation that chars are 1 byte, shorts are 2 bytes, and ints are 4 bytes, as FSX/FSUIPC is compiled. Is this correct?

Yes. Also see the thread entitled "About bits, numbers and hexadecimal ." in the FAQ subforum.

Also, NewWeather.h in the SDK is different than NewWeather.h in New Weather Interface for FS2004.zip. Which one is the correct one for FSX?

The later one. The only differences are in the new commands like NW_GLOBAL and the field for dynamic change rate, I think.

Pete