Good explanation of a traditional scanline rendering

A traditional scanline renderer begins by projecting the bounding box for each primitive onto the screen to determine the rectangle of pixels potentially covered by that primitive. A pointer to the primitive is stored in the first scanline in the bounding rectangle. Rendering takes place one scanline at a time, starting at the top. At each row, new primitives are added to an active Y list. The primitives in the Y list are then bucket-sorted in X. To render a row, the renderer marches from left-to-right, maintaining an active list of primitives as it goes. At each pixel any new primitives are added to the active list, which is maintained in Z-order. The pixel is then rendered by processing primitives in order from front-to-back.


이 짧은 글안에 traditional scanline renderer의 전체 프로세스를 정말  잘 정리해 논 글이기에 올려 본다.