Re: [Plib-users] culling problems w/ off-axis projection

[John F. F. <joh...@cy...>]

As you can see, I had a bit of trouble with the 
Windows e-mail program (Eudora) deciphering the 
Linux (I think) message, but I think I got it right.

I have put what I think is the correct change into SVN.  Please test it out.

         - John


At 08:45 AM 5/31/2010, you wrote:
>Sending again, because I used the wrong email 
>address. Twice! Sigh. * John F. Fay -- Monday 31 
>May 2010: > It uses the functions "_isnan"  and 
>"_isfinite" which are just > fine on my 
>Windows/MSVC platform. For some reason MICROS~1 
>is in love with these underscore functions, 
>which are, in the real world, just isnan() and 
>isfinite(). So this doesn't compile on 
>Linux:   sg.cxx: In function ‘void 
>sgQuatToAngleAxis(float*, float*, const 
>float*)’:   sg.cxx:1311: error: ‘_isnan’ 
>was not declared in this scope   sg.cxx:1311: 
>error: ‘_finite’ was not declared in this 
>scope I suggest to do something like this: diff 
>--git a/src/sg/sg.cxx b/src/sg/sg.cxx index 
>0e182fb..b6aa05c 100644 --- a/src/sg/sg.cxx +++ 
>b/src/sg/sg.cxx @@ -24,6 +24,12 @@ #include 
>"sg.h" +#ifdef UL_WIN32 +#  define isnan _isnan 
>+#  define finite _finite +#endif + + sgVec3 
>_sgGravity = { 0.0f, 0.0f, -9.8f } ; void 
>sgVectorProductVec3 ( sgVec3 dst, const sgVec3 
>a, const sgVec3 b ) @@ -1308,7 +1314,7 @@ void 
>sgQuatToAngleAxis ( SGfloat *angle, void 
>sgQuatToAngleAxis ( SGfloat *angle, sgVec3 axis, 
>const sgQuat src ) {    SGfloat a = (SGfloat) 
>acos ( src[SG_W] ) ; -  if ( _isnan(a) || 
>!_finite(a) ) { +  if ( isnan(a) || !finite(a) ) 
>{      ulSetError(UL_WARNING,"sgQuatToAngleAxis: 
>acos(%f) %f",src[SG_W],a);      if ( 
>src[SG_W] >= 1 ) a = 0;      else if ( src[SG_W] 
><= -1 ) a = SG_PI; m. 
>------------------------------------------------------------------------------ 
>_______________________________________________ 
>plib-users mailing list 
>pli...@li... 
>https://lists.sourceforge.net/lists/listinfo/plib-users