Questions about Shadow mapping

I have read the tutorial about Hardware shadow mapping from http://www.paulsprojects.net

I have also read the nvidia paper from GDC 2000:
“Shadow Mappingwith Today’s OpenGL Hardware”

And i have some questions:

  1. The depth map texture is projected on the scene to compare R with the deph stored in the texture. Ok.
    But if the texture is projected on the scene, Why is it not displayed in the color buffer ? (like in page 18 of the nvidia paper)
    I understand the depth map must not be displayed. But i don’t understand which function avoid it to be done.
    When i don’t use shadow mapping and when i project a texture on the scene, the texture is displayed on the color buffer. Where is the difference? The functions seems to be the same.

  2. How to display the projected depth map to the color buffer in order to have the same render we can see in page 18 of the nvidia paper.

3)GL_ALPHA_TEST is used to update on the screen the regions which are enlightened (not in shadow). If my scene has some semi-transparent models, could it produce some problems ?

Thanks.

I have understood why the projected texture is not visible in the color buffer with Dual Shadow mapping (because of the use of GL_COMBINE_EXT which keeps only the alpha value of the texture).

But for Hardware shadow mapping…?
Is it because the texture is created using GL_DEPTH_COMPONENT ?