>Is it possible to extend the due date by a day or two?  I seem to have
>underestimated the number of problems that I would encounter.  Ideas
>that seem trivial in theory become hard when trying to code in Matlab.

i'm worried that doing so would take time away from preparing for
finals in this and other classes, but ok:

+ projects will be accepted after thursday midnight until SUNDAY NOON.

+ if you turn in your project after thursday midnight, you MUST
  include in your electronic submission a jpg of the image.
  use IMWRITE; you might need to use a high quality setting to get
  a reasonable result.

  (if you turn in your project on time, a jpg would be appreciated
  but is not required.)

+ some kind of bonus points will be awarded for turning the project in
  on time.


here are some suggestions and remarks that you were intended to figure
out ahead of time or as you worked on the project:

+ if you work simultaneously on too many (interdependent) parts of
  your program, it can be especially hard to locate the source of
  problems.

+ ray-polygon intersection is required for (recursive) ray-tracing, 
  is independent of the kind of projection, and can be used to confirm
  the results of your item and light buffers.  thus, get it done early.

+ polygon meshes really slow things down.  a bounding shape, e.g. sphere, 
  should help a lot, as should item/light buffers.  rasterization
  (polygon fill) is pretty independent of the kind of projection.

+ every ray misses most of the polygons.

+ once a ray is shot, it is independent of the kind of projection.

+ you probably won't have that many spheres in your scene, and spheres
  are relatively fast compared to polygon meshes.  thus, if you're
  only going to get the item buffer for either spheres or polygons,
  you should get it to work for polygons.
-- 
Thomas Yan