> As far as you question is concerned, look into the 3.1 source code, it
> does not use those caches and is probably easier to understand. 
> Either way, the light code will be replaced sooner rather than later
> anyway, so looking at the clearer 3.1 code won't hurt much.

Yes, you are correct that it is easier to understand. It also doesn't
have lightgroups, so it is easier to see what is going on as a whole.

However, I am still puzzled about update_light_list_entry. In it, about
80% of the way down (I can't give you the exact line number, sorry)
there is a line:
//    if (insert)

Why was this if commented out? It seems to me as if it should not be.
With it left in, if the ray fails the intersection tests above, the
light source is left as inactive. Without it, the light source is always
marked as active, even if it does not affect the ray, and both s0 and s1
are marked as zero.

Having it commented out would seem to me to give the same results but
a performance decrease, so does anyone know why it was done? Does it fix
some problem I haven't thought of?

Thanks again for any help you can give,
Daniel

-- 
  ,^--^.      ,-----.      Never meddle in the affairs of angry cats,
 ( +  + )-----   ---'      for they are well-armed and quick to bite.
 /  --          )       me> http://surreal.istic.org/ key> 885b170d
|_,-|_/--,_|-\_|     http://www.xcom2002.com/doh/index.php?s=04113018oth

Post a reply to this message