What are the plans regarding NodeMaterial?

Over the years a common request from users has been the ability to modify built-in materials.

Built-in materials do a lot of stuff, but sometimes you only want to tweak a couple of lines of a giant shader, or inject some GLSL at some stage.

There has been a plethora of PR’s and issues posted about this. Some are open, some have been closed:

#14232 #14231 #14206 #14166 #14099 #14098 #14031 #13198 #10791 #13446 #14009 #14011 #13192 #7581 #13364 #12977 #11562

At some point between late 2015 and today, this PR #7522 has been quoted as a blocker for some of these, if not all of these.

As is, #7522 is just another example from /example, until it’s not moved to /src it shouldn’t be treated as a first-class citizen. Why is an example (a very large one at that) blocking small PRs?

I see this as a problem, and I wonder if others do too.

[mrdoob]

Sorry for the confusion…

The fact that NodeMaterial is “blocking” all these PRs is because NodeMaterial feels like a good design (modular, tree-shackeable, extensible, serialisable, …) design for materials.

The design of the current material system is hacky and hard to maintain. I would rather not add things that make it even harder to maintain on top. I would, instead, focus on exploring NodeMaterial.

[mrdoob]

@pailhead do you mind renaming the title of this issue to something more descriptive?