I thought the current developer was involved with the origins of dyntopo way back in the day.
2 Likes
Thanks for the kind words. I think sculpt mode will attract more developers once we do more code refactors. On the topic of multires attributes like paint or face sets, that’s actually more a problem of getting all the relevant developers to sign off on a design. Code-wise it would be fairly simple to implement (Pablo de Barro did actually have a vertex colors implementation for multires).
20 Likes
I was involved in the “new” edit mode 10 years ago (that was back when Blender only supported triangles and quads). Geoffry Bantle wrote the core data structure (BMesh); I took that and rewrote all the old edit mode tools to use it. I also wrote the knife tool.
BMesh was subsequently used by Nicholas Bishop to code Dyntopo. BMesh has also been used for the bevel modifier (I believe several people contributed to that) as well as booleans.
At this point I’ve rewritten most of Nicholas’s code but the core design concepts are still there. His approach to subdivision ultimately proved to be ideal; the rather large amount of time I spent the past 2 years trying to improve upon it ultimately went nowhere.
21 Likes
When you say problem you mean there’s a debate and some developers aren’t ok with the design? or that design doesn’t exist yet and needs to be fleshed out?
I have noticed, following development last year, that usually that’s the case, design discussions are much larger portion of feature implementation than coding. By far. But so far I haven’t seen a design choice I didn’t agree with so it’s worth the wait I guess.
The design doesn’t exist.
are you having performance problems with the viewport of blender 3.6?
It’s choking a lot
Thanks for the explanation. I know you must be very busy with work on sculpt mode for Blender.
Great to know that multures is simple to implement. Thats a huge relief.
Is there a reason why a design doesn’t exist? Feel free not to answer this especially if you aren’t supposed to discuss this stuff outside inhouse discussions. I am hoping you guys come together and hash out a design for multires hopefully soon. I remember Pablo implementing the vertex color for multires sometime back.
More power to you. You are doing tremendous work on sculpt mode for Blender.
2 Likes
It’s because there are several ways we could do multires attributes. We could have a special multires domain, or we could pack the other attribute types into multires. There’s also the question of whether or not to support standard PTex, and if so should it be the sole way to paint attributes into multires or should we have another system as well.
12 Likes
the one thing what we really need is shape key support for multi res. thinking of HD expressions 
4 Likes
Attribute paint support! I would like to use Multires in my workflow but i just can’t. Sad.
7 Likes
Multires is such an essential feature. I hope it will be improved for 4.0.
9 Likes
Thanks for the info.
I would say adding Ptex would be a nice move but also adding more options if possible is always a good thing to give users more options if they find the other method more suitable in their workflow.
But I think Ptex would be a very good option especially if you can bake to uvs later. Ptex will not require as much subdivision as vertex paint would require giving better performance.
Development is still being made by Disney on github: https://github.com/wdas/ptex 
EDIT: On second thought, I think vertex paint might be better as you can use sculpt vertex paint with face sets and masks. I don’t think that is possible with ptex (you can’t use it with face sets or masks while painting). Sculpt vertex colors also allows you to have slots/layers in the color attribute tab so I think ptex might be a whole workload compared to building on the system we already have with dynotopo for multires which is color attributes/sculpt vertex color.
1 Like
Julien created design task about ‘UV Map from Face Sets’ function that we came up with in this thread. And Hans added Face Set and Set Face Set nodes for Geometry Nodes. So both face set generic attribute and UV Unwrap coming in 4.0 
I’m not sure yet but if face sets can be updated real-time in geometry nodes, that opens up door to a whole new creative functions we can come up with. Very exciting feature for Sculpt module
15 Likes
@nickberckley What are the use cases to convert face sets to vertex groups?
For weight painting this doesn’t seem useful since the vertex groups are all overlapping.
Also people requested to convert face sets to material slots. I also don’t know why yet.
Simply for consistency of selection sets.
We don’t have operator for Face Sets selections in edit mode, in Paint masks Face or Vertex selection modes.
But in those modes, we have buttons to select vertices or faces according to vertex groups or material slots.
Selections for addons are also based on that.
So, if you don’t want to constantly go into sculpt mode to hide mesh according to face set to change selection in those modes, you have to convert face sets into data that have an operator for selections.
(In case, UVmap wanted is not supposed to have UVseams corresponding to face sets boundaries.)
If we would have masks/selection, defined as in future paint mode design and accessible in all modes, request would not exist.
But until this achievement, that is normal that request of conversion of Face Sets into material IDs or Vertex Groups will be recurrent.
4 Likes
It’s for the ZRemesher QuadRemesher addon, it uses material slots to guide the remeshing process.
4 Likes