unsigned long long for lat/long?

[TurbofanDude]

HI there,

I am trying to read the latitude and longitude from FSX. I have gotten the raw, unformatted data from FSX, but I am having trouble converting it to a usable format.

I have converted the unsigned long long to a float while applying the conversions, but I am either getting inaccurate data (i.e. 42.7445 when it should be 42.4542) or I get just even 42.

Here is my code: (There is a call to FSUIPC_Process in there, I just haven't actually included it in the example).

signed long long LAT;

FSUIPC_Read(0x0560,8,&LAT,&result);

float y = LAT;

y *= 90;

y /= 10001750;
y /= 65536;
y /= 65536;

p->latitude = y; //p is a user defined an aircraft data struct[/CODE]

Thanks in advance,

--

Collin Biedenkapp

[Graham Pollitt]

In vb.net I use 'long' variable type then copy it into a 'single'.

'code below is snippets from various parts of a class module in vb.net'

Private fsLatitude As Offset(Of Long) = New FSUIPC.Offset(Of Long)("position", &H560)
Public CurrentLatitude As Single

FSUIPCConnection.Process("position")

Me.CurrentLatitude = Me.fsLatitude.Value
Me.CurrentLatitude = Me.CurrentLatitude* 90D / (10001750D * 65536D * 65536D)
[/CODE]

What language are you using? Why 'long long' and not simply 'long?'.

[TurbofanDude]

I use long long because I need a 64 bit integer, where as long is 32, int is 16, etc. It's a whole mess in C++ that isn't needed in .Net.

I got a decimal at least, but it's still wrong (using a method similar to yours). I'll have to review it in the morning.

[GHD]

You are losing precision by copying to a float. Use a double.

long long Lat;
double dLat;

if (FSConnection.ReadAndProcess(0x568, 8, &Lat)
  {
	dLat = Lat;
	dLat = dLat*90.0/(10001750.0 * 65536.0 * 65536.0);
}
...
...
[/CODE]

[TurbofanDude]

I did the above conversion, I got 42.741640579659617 when the in-sim latitude read N42*44.50'

Is that normal?

--

Collin

[Pete Dowson]

I did the above conversion, I got 42.741640579659617 when the in-sim latitude read N42*44.50'

Is that normal?

By "normal" do you mean "correct"? If so, of course it is correct.

42.7416406 (the nearest I can get in my calculator) = 42*44.498436 which would certainly be 42*44.50' to the nearest 1/100th of a minute! What is your problem? Check it yourself on a calculator (just multiply the fractional degrees by 60 to get minutes of course)!

Pete

[TurbofanDude]

'just multiply the fractional degrees by 60 to get minutes of course' - that seems to have been my problem. It slipped my mind that we are dealing with fractions of 60, not of 100, therefore, I was reading it as 42* 74'15, without multiplying to get minutes and seconds.

Thank you for the help everyone, as it is working now. This should hopefully have been my last stupid question.

--

Collin