|
![](/i/fill.gif) |
On Fri, 21 May 1999 16:15:12 -0500, Christian Wietholt wrote:
>In the makefile there is a little error. I think the two
>lines beginning with C:\.... are there accidently.
Yes, they are. cpp -MMD is supposed to ignore system headers, but
apparently it missed some (I'm using egcs 1.1). The dependencies
were generated automatically and massaged by a Perl script, so
while they're accurate, it's easy for bugs like that to sneak in.
>Second:
>
>file f_func.c line 207
>
> //*(calc_stack_top)= pow(
>ABS(*(calc_stack_top)),*(calc_stack_top+1));
You are correct. I had to fix that one, too, but I forgot to add
it to my zip file.
>Third:
>
>Here I don't know exactly what to do.
>
>"../f_param.c", line 78: error(1164): argument of type "OBJECT *" is
> incompatible with parameter of type "PARAMETRIC *"
...
>I put here a cast in like :
>
> if (! Intersect_Par_Box (Ray, (PARAMETRIC*)Object, &Depth1,
>&Depth2, &Side1, &Side2))
I'm sure this is okay. I'm surprised my egcs didn't complain about it.
OBJECT is the generic superclass to PARAMETRIC. You got to that line
through the POV equivalent of a virtual function pointer, so you know
Object is in fact a PARAMETRIC. The cast should be completely safe.
>PRECOMP_PAR_DATA *Precompute_Parametric_Values (Par, flags, depth)
>PARAMETRIC *Par;
>char flags;
>int depth;
>
>to
>
>PRECOMP_PAR_DATA *Precompute_Parametric_Values (PARAMETRIC * Par, char
>flags, int depth)
This should be fine, too. I need to go eradicate the last vestiges of
K&R from all this code someday; does anyone know a warning that I can
enable to point out K&R-style function headers?
Post a reply to this message
|
![](/i/fill.gif) |