OK, well obviously I can't really argue regarding the programing difficulty, integration or whatever, but honestly this answer (or the longer version for Forrest Pro) is pretty... I'm trying to think of a diplomatic way of saying it, but can't.
You're basically saying that the product that we've been using and come to depend on has now lost a
major compatibility component, and you're not in any hurry to fix it? Really?
So, no support for Mental Ray,
because AD dropped it, yet Mental Ray for 2018
is already available and is slated to have even more development than it previously did. Students or newer users who may have been renderer agnostic can jump around to whatever they feel like, but many pros will be buying the MR Pro license and are already using the free version now. Why? Workflow and legacy file scenes for starters.
Have you considered the complexity on
our end, where we may have hundreds of scenes, and thousands of objects built up over time utilizing Mental Ray exclusively? I
have VRay and find it lacking in any number of areas, but again, that's not the biggest issue: long-time users of one system or another depend on the libraries they've built up over that time. You could tell me that Pixar was
giving away Renderman for 3DS Max for
free and I still would be switching outright. Maybe over the course of a couple of years, but in the meantime I have a lot of work to do. Clients don't care "how" it gets done, just what it looks like and how much it costs. There's no room for "Well, I know I did it that way before and for this much, but it's going take an extra couple weeks and cost and extra 25% because I need to switch my workflow over to...
How well does anyone think that'll go over? At the very least, you might want to survey your users before assuming that the majority of users either only using VRay or don't care what they use. Along with that, is there any workaround to use in a Mental Ray workflow. So far I've been building with Railclone in max 2017 and then collapsing the meshes for export to Max 2018. Obviously not ideal as instancing and many other features get sacrificed using this method.