Yes, that’s true, as an option it would be the way to go
That shouldn’t discard the argument that Blender would be well served with the modifiers moving to a nodal system and leaving the stack-based system behind (especially if we can have more granular modifier types on top of what we have now). I do agree by the way that the current way of modeling shouldn’t be completely abandoned.
Agreed, I hate modeling in houdini.
I know what you mean - although I’d argue missing functionality in the Nodes isn’t a problem with the Editor itself but Cycles Material System. The things I’m missing from Groups - creating boolean inputs, exposing color ramps in Node Groups and including image files in the Node itself.
I don’t think that is very comfortable.
I think the stack should stay with the additional functionality that it creates the nodes for you.
The reason i am thinking this is because i don’t think nodal modifiers are needed in more than 25% of all cases. Maybe even less.
More often than not, you just want to throw a mirror, solidify or subD modifier on top of an object and having to open a node editor windows for this action every time is not very comfortable.
It only makes sense if you want to create a procedural model or keep parts of it non-destructive.
For standard destructive modelling it is way to slow, but other than that its a pretty awesome modeler with all of its functionality. You can do a lot of stuff you can’t do with any other program.
Agreed with that. Maybe there could be a mode switch between classic linear modifiers and node-based modifiers, selecting the former would show modifiers in a list like it does now but internally it would use whatever code an equivalent node tree would output. I guess we could have the best of both worlds. However showing an actual node tree in a list is difficult, the main problem I see is how to show branching ? What if a “subsurf node” output goes into a “bevel node” and into an “extrude node” ? Which one do you show in the list after the subsurf ? Cycles nodes do this with a plus sign that’s collapsed by default and can be expanded, but it’s kinda hard to navigate.
Ok we should definitely start a wishlist. I want node groups !
No need for a mode switch. If you use the stack (for creating modifiers) you’ll get the usual linear top down hierarchy, which in the background creates the nodes in the same fashion (in the node editor). If you want to branch you’ll need to use the node editor.
The behavior could be similar to materials in the property editor. You still could change the modifier data in the property editor, but if you want to change the way nodes are wired together you’ll need the node editor.
Best of both worlds, comfortable for fast and complex workflow.
What do you mean?
Like spherical or tube mapping. If you want this on procedural texture generators you have to build quite a complex node group to do it.
I wonder if it will be possible to make a scene graph like Guerilla, it’s really powerfull.
A bit of a speculative subject, but just wanted to comment regarding modeling. People underestimate how much work it takes to do lowlevel nodal setups. Modifiers are incomparably easier and faster to use. The best solution would be 3dsmax like MCG graph where modifier system is unchanged, coexisting with new types of modifiers that can be defined with graphs( e.g “NodalModifier” type with reference to a graph).
Still as we can see with other nodal systems like ICE and MCG, it takes many years to populate them to give users the control they seek vs what they have with .py. In contrast with Houdini we see users preferring VEX(code) over VOP(nodal) as it’s just a lot faster and straightforward to work with. I hope modeling operators (such as boolean) are exposed so you could do things like InputA,B>Boolean->Detect edge->process edge/rebool result>output.
I agree for most part. But it is also a mindset and depending on task. Using Grasshopper for Rhino often it is ridiculously easier to use nodal setups compared to code due to the visual feedback. Connect a node and see the result. Connect another and compare, connect two and experiment.
Also when working with many to many relationships the “lacing” nature of GH is tremendously intuitive.
would Everything nodes have included any kind of particle solver similar to the addon MOLECULAR? it was a good aproach, and it could be implemented with new relations between the particles as for example “disolution”. It would be usefull for the creation of addons that mimic geological erosion but in trully volumetric way.
I’m wondering if Everything Nodes will make Animation Nodes to be obsolete. What do you guys think?
Well, the title of feature says it all. Everything as nodes.
So, of course when goal will be reached animation nodes addon will be obsolete.
But that is a long term todo. Everything will not be accessible as a node in 2.8, first beta.
It is the same developer who is making “everything nodes” and made animation nodes.
Probably, he will not have time to make an update of animation nodes for 2.80.
So, there is a probability to have “nothing nodes” in first 2.8x releases.
But when feature will be there, community will share nodegroups, demos and create addons for it.
So, instead of 2 or 3 popular custom nodes addons, 2.8 will probably have a dozen of popular addons made for everything nodes.
Modeling on nodes is good. Example Modo
Yeah, that is what I’m wondering. It makes me think of not diving right now into AN and wait a little longer.
Any news on when Jacques will return to the Blender institute?
Exciting times ahead.
Wohoo, can’t wait!