Nodes everywhere!!!!

I watched a video of a new plugin of 3dsMax, Ephere Lab. I thing that can be interesting to see the workflow, specially because looks this guys are focus in easy plugin development. We need nodes everywhere!!! and a cycles bake textures…:slight_smile: Video: Have fun!

Meh. Houndini realesed somethink much more intresting.

Sooooo… ICE?

See number #20 (second column to the right) on this list:

I saw a post on here where someone had an ICE like system they were developing for Blender btw. Dont know where thats at but it looked impressive. I dont think the developer was collaborating with the BF devs though.

there is particle nodes… which has been in the works for a while –

he is actually working for the blender institute now… weekly updates are here –

+9999 interwebs. My hopes are with Lukas Tönne now being hired by the foundation, I hope he´s not just sitting bugfixing (I know it´s an essential part, especially by Tons tweets. But Lukas had some really interesting thing going on his blog phonybone with the progress of nodification of everything)

If you read the link i posted above, in the first week he actually mentions it

Refine the proposal for particle nodes roadmap. First rough outline here.

and again

Make a proposal for the first particle overhaul stage (buffer system, attributes, point cache interaction, duplis). Most of it exists on paper, needs to go to wiki. Implementation should be pretty straight forward, especially if good separation of concerns can be maintained (e.g. keep particles agnostic of point cache details and dupli system).

I’m pretty sure Pynodes can already be used to make a system just like this. There were some examples that were posted before that do exactly the same type of stuff. The only reason we don’t have a system like that now is that someone needs to write all the nodes. :wink: But yeah, that’s the idea behind Pynodes at least. A generic interface to create your own nodes in python.

I don’t think pynodes is completely finished and up to the task of using nodes for everything just yet. Evaluation problems withdependency graph come to mind for example…I’m sure there are others.

All of the things needed for a system like ICE aren’t exposed to the Python API. On top of that python is slooooow and hard to scale with processor cores/threads. Part of what makes ICE so attractive is that almost every node is fully threaded and the modifiers you create with it can actually be FASTER than the built-in compiled versions.

I like a lot of things about Blender,but if I have to be honest there is a thing I hate,python.
A scripting language is a good thing for small tasks but in Blender it’s more a way to extend the program,where nodes or a plugin system should be used instead(for me python is good only for some modelling function,even exporting I think is arguable).
The only Blender weak point is this,it’s hard to extend,even if in the community there are a lot of talented people that could extend the program (with the right conditions) now all the work is done by blender coders,development it’s too centralized.

Moved from “Latest News” to “Blender and CG Discussions”

Python isn’t the problem at all, it’s the API. Python is being used as a platform for lots of high-performance computing, because there are many ways to bind native code to it, while the code that doesn’t need to be native can stay Python.

If the Blender Python API provided efficient, direct access to mesh and image data, it would be sufficient.