Quote -
-Monte Carlo Method.
-Metropolis Light Transport (w/ multiple intensity variant algorithms). <- kind of the same as the MC, but with some differences.
-Contrast controls (temporal and spatial).
-Multi-Pass Rendering.
-CAT Image Compositing.
-Selective Material Overrides.
-Mental Ray Connection Shaders on all material. <- you can omit that one, if you insist the renderer specific shaders are irrelevant.
-Material and Object based Photon Emission.
-Global Displacements.
-Volumetric Refractions.
-Final Gathering.
-Scripted Pre and Post Rendering operations.
There's more too, I'll see what you have to say about why none those have any direct impact on the rendering process compared to a raytracer that doesn't have them first.

What do any of those features have to do with raytracing in general, which is the only thing I was referring to in the other thread?  Multi-pass rendering?  Global displacements?  None of these things have anything to do with either thread in question.  I never said Mental Ray couldn't do all these things, or argued that it wasn't a great renderer for those kinds of advanced features.

All I'm saying is that the process of raytracing IS general for the most part, and the same basic principals make up the process for every raytrace engine out there.  That is to say, raytracing is a process of following rays from a virtual eye or camera point OUTWARD, rather than originating at the light source.  It works by tracing a path from the imaginary eye through each pixel in a virtual screen, attributing the contribution of each light in the scene to that pixel.  As each ray is cast from the eye, it is tested for intersection versus objects within the scene. In the event of a collision, the pixel's values are updated, and the ray is either recast or terminated based on material properties and maximum recursion allowed.

So what are you in disagreement of again?