SDK Question VESSELSTATUS2 surf_lat/surf_lng return weird results

[Bj]

Using this;

DLLCLBK void opcPostStep(double simt, double simdt, double mjd)
{
    OBJHANDLE oh = oapiGetFocusObject();
    VESSEL * v = oapiGetVesselInterface(oh);
    VESSELSTATUS2 tmp;

    v->GetStatusEx(&tmp);

    sprintf (oapiDebugString(), "%f, %f", tmp.surf_lat,tmp.surf_lng);
}

with the API ref saying;

surf_lng longitude: vessel position in equatorial coordinates of rbody [rad]
surf_lat latitude: vessel position in equatorial coordinates of rbody [rad]

Shouldn't the results be...?
[math]-\pi < x > \pi[/math]

Why am I getting insane numbers like 1 trillion+?

 

[dbeachy1]

That's because you're telling sprintf to parse two 64-bit doubles as two 32-bit floats. What you need is this:

sprintf (oapiDebugString(), "[COLOR="Red"]%lf[/COLOR], [COLOR="red"]%lf[/COLOR]", tmp.surf_lat,tmp.surf_lng);

 

[computerex]

I think you didn't initialize the VESSELSTATUS2 structure correctly.
Following test code produces expected results:

#include <stdio.h>

int main()
{
    char buffer[256];
    double PI = 3.14159265;
    sprintf(buffer, "%f", PI);
    printf("%s\n", buffer);
    scanf(" ");
    return 0;
}

VESSELSTATUS2 should be initialized with:

VESSELSTATUS2 vs;
memset (&vs, 0, sizeof(vs));
vs.version = 2;

 

[Enjo]

Yup,

• The version field of the VESSELSTATUSx structure must be set by the caller
prior to calling the method, to tell Orbiter which interface version is required.
• In addition, the caller must set the VS_FUELLIST, VS_THRUSTLIST and
VS_DOCKINFOLIST bits in the flag field, if the corresponding lists are
required. Otherwise Orbiter will not produce these lists.

Nevertheless, you really should use "%lf" on doubles, to make your code compiler-independent.

 

[computerex]

#include <orbitersdk.h>

DLLCLBK void opcPreStep(double simt, double simdt, double mjd)
{
	VESSEL * v = oapiGetFocusInterface();
        VESSELSTATUS2 tmp;
	memset (&tmp, 0, sizeof(tmp));
	tmp.version = 2;
        v->GetStatusEx(&tmp);
        sprintf (oapiDebugString(), "%f, %f", tmp.surf_lat,tmp.surf_lng);
}

Seems to work

 

[Enjo]

Yes, but please don't place code with errors like "%f" on doubles, which is supposed to be an example code. Others may copy it and not realise why it doesn't work on their compiler, you know.

 

[computerex]

Yes, but please don't place code with errors like "%f" on doubles, which is supposed to be an example code. Others may copy it and not realise why it doesn't work on their compiler, you know.

Meh, I like using %f

ps. You realize you are more likely going to come across a compiler that doesn't know what "lf" is then one that won't properly do the conversion of doubles with "f", right? And Orbiter libraries are only compatible with Microsoft's compiler series.

---------- Post added at 05:20 AM ---------- Previous post was at 05:07 AM ----------

BTW it turns out that "lf" isn't even a valid input for sprintf when giving doubles as arguments...

 

[Enjo]

There's no excuse for using "MS C/C++". Sometimes a programmer has to go outside of the Orbiter, and even the Windows world.

Anyway it seems that you're right about that "lf"

http://www.cplusplus.com/reference/clibrary/cstdio/printf/

The lowercase l applies only to integer and character types. There's also capital L, but it applies to "long double"

I swear it was "lf" for double back in the Old Days.

[EDIT]

If you think about it, the capital L instead of lower l for long double may be intentional, so that old code which used lower case l for doubles still compiles properly.

 

[Bj]

Hah your awesome guys thanks a billion

 

[dbeachy1]

I swear it was "lf" for double back in the Old Days.

Like me, you are probably thinking about scanf, which does require "%lf" to correctly parse a double and "%f" to correctly parse a float. After some research, it turns out that the compiler does indeed auto-promote floats to doubles before passing them to functions that take variable arguments. In other words, you can get away with "%f" for doubles for printf, but not for scanf. More info here: http://stackoverflow.com/questions/210590/why-does-scanf-need-lf-for-doubles-when-printf-is-okay-with-just-f