Hi All , I would like to point your attention to this oss sw…It’s a sw agnostic node system (ala ICE) … which would be easly integrated into blender:
http://coral-app.com/
http://code.google.com/p/coral-repo/
Video !
Is anyone willing to code a modifier to get it integrate in blender ?
(check how is already integrated into maya, and how powerful is !)
Cheers 
L.
looks nice
atmind: Officially started working on the #OpenCL implementation of the #b3d #compositor.
http://t.co/vzqjZRVx
regarding nodes jbakker is starting on opencl-ing blenders nodes! ^^ coral looks nice but why use only CPU and only intel cores.
interesting indeed , but more powerful of the computation is the flexibility from my pov , and coral permits a lot of flexibility (sw agnostic , pluggable to any sw)
and it can be always moved to gpu…
L.
Bakker’s work is very, very different to what this is.
This is like node based modifier stack, where as Bakker (ATMind I believe?) is converting the ‘compositing’ nodes to OpenCL.
This could be forwarded to Lukas Tonne who I believe was working on something like this for the particle system and mentioned (if I remember correctly) something along the lines of this type of set-up.
Whether or not it could be easily integrated is another thing, but, and it’s a big but, there were some way to have external solvers and things of that nature in Blender which could be used with tools like this then that’d be a very awesome addition for the TD’s people out there.
it could be like grasshopper is for rhino, and that means a really huge “plugin”!
I wonder if you could turn this into an open source type version of SLIM(Renderman node based shader generator), it could be really useful in that sense.
it shouldn’t hard to connect it to a shading system , but having a chance to get it connected to blender it would provide the ability to bring new functionalities (spline rig anyone?) to blender , or any other sw (maya/xsi/max…), and share the results…
oh btw , is there any plan for having a plugin system for blender ?
(even just for modifiers…)
I was also reminded about Lukas (phonybone) work:
http://phonybone.planetblender.org/2010/08/23/one-two-many
but Coral does look very interesting. Wonder if some sort of compatibility between Coral and what Lukas is creating could be done?
@tyrant monkey
Have you seen Shrimp?
http://shrimp.sourceforge.net
Great project, Blender should take this really seriously.
But(as always),projects coming from the outside of the Blender community are not taken into the right consideration…
@san I have known about shrimp for a while now but it never hurts to have more of these kinds of editors.
but isn’t coral just the engine behind the nodes? so let say the video shows transform nodes, loc,rot etc. but the functionallity must exist in blender?
so lots of the nodes we saw in the maya integration is surely using maya tech, not coral. coral is jut the GUI. so… having them in blender would just have nodes we could drop in an connect but it wouldn’t do anything to the mesh actually. or?
it sure looks good but blender already has a node system written, gui and everything. and when phonybone ha re-written it to more flexible so maybe even addon developers in python can create nodes etc we’ll be there with blender nodes for almost everything.
There’s been a lot of recent work on Blender’s build in node engine, I’m not sure if it would be a recommended step to stop all that work, then rip all that code out and replace it with an interface using the coral engine.
For one thing it would take quite a bit of work to remove the existing code, hook up the coral engine, and rewrite all the nodes using C++ as the coral engine uses it, I’m not exactly sure there would be a net productivity increase in the total amount of node-types and node-based features in Blender for most of the near-future if that was to be done.
Dataflow system ( like ICE in XSI , Houdini or Coral) is very powerfull and it’s the future for particles , deformation or anykind of vfx.
Hope PhonyBone project for Blender continue and reach the flexibility and the power of others systems (with availability of alot of node type).
This is amazing - it should greatly increase blender’s capabilities.
dataflow system is more “technical artist” friendly than programming. Many 3d software actually take this road. And the node is a C++ librairy , I think is faster then programming in python with the advantage to use multithreading and in the future … gpu ?
And after that artist can share is nodes setup with the community, where a repository of modifier/particles setup can be made. As you said kbot , I think it should increase blender capabilities. It’s the future and Blender must be there !