I thought they added or fixed this so you could adjust custom props even when items are linked
I use Blender 3.3 latest. The Lodify addon did not detect switching between solid, shading, or rendered views. I have to click on the “Level of Detail” buttons every time after switching the viewport modes. Will there be an update for this addon,
Cheers!
To be extra clear
This is lodify
This is what a proper LOD or PROXY system would look like
I tried hard to first implement LOD/PROXY
first via python API
then via geometry node + python
i can guarantee that we are all wasting our time on this subject
a proper implementation should be done in blender C++ directly
And you have proven this point by trying to get it to work through alternative means.
@BD3D could developers of addons that benefit from the implementation (Scatter, E-Cycles, K-Cycles, Grasswald, Octane, etc. ) co-fund a third party developer who could get the C++ implementation done? So basically crowdfunding the feature. I highly doubt that any of the core developers have this on their to-do list for the immediate future.
I believe the blender “properties > object > display as” needs a lot of new features.
ex: display as solid bounding box, display as solid convex hull, display as point cloud, display as proxy with a solid memory management (option to only load textures & geometry on final render for example)
My company could fund between 5 to 10k $ on the project in the near future
we could start a fundraiser on the idea, i know large archviz studios that can’t switch from 3dsmax to blender because of this crucial lack of feature
But few things
-
It would be about display solutions, LOD is another subject
-
Unsure that feature-oriented fundraisers are accepted? If I got confirmation from the blender dev management team that improving blender display and feature specific fundraising is a welcome idea, I could start the initiative & contact potential backer.
-
What about a non-rasterization visualization solution instead? Rasterization is the real problem here. there are better viewport solutions for large polycount
Agreed, that’s where it should be.
To my understanding, they should be accepted. But it is important to start a design document first, make sure the developers who need to review the work schedule in time (so the reviewing won’t let the patch linger in limbo). And I think it would help to also lay out a proposed UI and a simplified explanation of the technical implementation before actually funding nad starting the development work.
contacted the dev on the blenderchat
*edit
and it doesn’t look good
I wished the foundation was more open to a more “decentralized” development process let say…
I think the problem here is not the money, but actually finding a developer that has a good enough understanding of Blender’s core architecture to implement such a feature. You would basically need hire @sergey or someone similar to get this done properly
I think the problem here is not the money, but actually finding a developer that has a good enough understanding of Blender’s core architecture to implement such a feature.
Yes it looks like it,
even if we make a task that get approval,
and even if we secure a certain amount of funds,
finding the right dev will be complicated, it appears that this is not a surface-level implementation
Is there anything else they mentioned? Or would the funding allow to get a third party developer to do some work for someone like Sergey so he can look into this? Basically indirectly funding the feature…