are the different covered pixel color used to signified that they belong to different (adjacent) triangles?
I found the link to the source mentioned in the slide:
This falls under the "Triangle Rasterization Rules (Without Multisampling)" section:
"The light and dark gray covering of the pixels show them as groups of the pixels to indicate which triangle they are inside."
So, in the bottom left group of 3 triangles, pixels in the leftmost triangle are dark gray, pixels in the middle triangle are light gray, and pixels in the rightmost triangle are light gray.
I find it interesting that for the top right group of 2 triangles, the lower triangle does not "cover" any pixels and is this not colored. This shows how supersampling by averaging out sample values in a pixel can help prevent aliasing.
Hope this is correct and helps!
Does the top edge need to be perfectly horizontal? i.e. both it's vertices/endpoints need to have the exact same y value for it to be considered a top edge?