Viewport FX III

The first major set of changes done as part of this project is now under review
https://developer.blender.org/D643

This for example will allow for multiple backends for graphics in the BGE (including Direct3D and AMD’s new Mantel API), this could also pave the way for BGE games or Blender on Android.

good! i prefer such developers who stick to the stuff then developers from foundation after few actions… in last month.

Any news about new thing? I saw that “something” work on linux now but don’t know how far it is from new viewport.

He’s still working on his project, most of it by now is preparing the code for review so he can do what is needed for the final merge into Master (review can be very difficult if it’s all submitted in one massive chunk).
http://lists.blender.org/pipermail/soc-2014-dev/2014-August/000196.html

I need to get things submitted for review purposes, but this year I’m planning on continuing the submission/review process after SoC, and not losing momentum like before.
Thats great. Most important information for us. :slight_smile: If he show some working beta builds I can donate this project. :slight_smile:

Every year, the last weeks of Gsoc, students say: “It´s not 100% ready but I will work on it after Gsoc”. Color me skeptical. I do hope that I´m wrong…

Donating for this project seems to me a bit odd,to be honest.
It’s weird,there are some projects that are discarded from gsoc because they are too ambitious while at the same time we have this,a project that will never end(I hope that I’m wrong but I’m a bit skeptical ).

It’s ringing all sorts of alarm bells for me too. Jason has stated this after pretty much every GSoC, but the last thing I remember actually being merged from them was bits and pieces of the sculpting stuff way back when. Maybe the merges just flew under my radar.

The patch that’s under review is just context creation, it’s complicated, but won’t make Blender’s viewport either pretty or fast by itself.

a new bmesh like case?
I have very few request and needs in blender:

  • edit two or more object at some time
  • a outliner like cinema 4d
  • a rock solid viewport ala softimage

get in line

Next Generation OpenGL Initiative - Call for Participation

So by the time this project could lead to performance improvements or nice viewport features there might be better API that could give more with less lines of code.

Oh god, yes to all three.

Make that Maya’s viewport 2.0 and you have my agreement.

From my understanding this project has a wrapper layer between the graphics stack that should make it easier to use different graphics backend such as direct 3d or even this newer api in the future.
Some one please correct me if I’m wrong.

Sure, but I think is bit much for now. If we can have a viewport capable to handle large scenes and maybe very dense meshes, will be good enough for me (and I guess for many others)

Agree, I don’t think having advanced shading is going to do much unless the viewport gets much faster to begin with (and doing it all at once may be too large of a project, hence why the Viewport FX project is just to rewrite the foundation).

As far as I know the core of Viewport FX is about done, which is why Jason is now concentrating on doing what is needed to prepare the code for final review and then do any needed changed for the final merge (a good chunk of it now at that stage such as seen here, here, and here).

thanks for the news! Any ETA?

Forgive my ignorance. Isn’t “ViewportFX core” just new method of binding OpenGL or OpenGL ES to blender? I don’t have much knowledge how it works but for me it looks like just preparation for actual viewport code. Sounds like two more GSoC to see and feel real new viewport in blender. :wink: Am I wrong? Is creating new viewport after this core stuff will be as easy as creating one advanced GLSL shader? Can someone elaborate on this?

Viewport 2.0 isnt just about advanced real time rendering or custom shaders, its a performance beast. One doesnt have to sacrifice the other. It just puts far more control in the hands of the user at the same time.

Don’t think Direct3D or Mantle API are on the roadmap, there is a reference to Direct3D in the patch - only because of the Angle OpenGL API which can do OpenGL ES via DirectX, - https://code.google.com/p/angleproject