pyOpenSubdivision

A Python wrapper (implemented using ctypes) for the OpenSubdiv C++ Library far topology refiner.

Updates

• Added keyword arguments (verbose (-v) and level (-l <N>)) to ctypes_subdivider.cpp (these are handled by subdivider::settings).

• ctypes_subdivider.cpp handles level 0 subdivision better now.

  • Within the refine_topology function, an assertion is made that the number of vertices per face of the refined mesh is 4, i.e. it consists entirely of quads.

  assert(fverts.size() == 4);

  This assertion is only strictly true if the mesh is actually subdivided (maxlevel > 0). Otherwise, it is possible to feed a mesh containing ngons, which if not subdivided (evaluated for maxlevel = 0), will fail the all quads assertion, crashing the code.

  It would seem reasonable to simply ignore the all-quads assertion, but the block that constructs the edges (for (int i = 0; i < nn_faces; i++) { ... }) for the refined topology relies on the assumption. Furthermore, ignoring the all-quads assertion would involve returning (to python) a set of faces where each could contain an arbitrary number of vertices, which is very difficult (to the extent that I could not figure out a solution). Conceivably, it would be possible to use the vertsPerFace data to reconstruct a flattened (1D) faceVerts list; but ultimately, any solution tends to involve reconstructing information that was passed into the refiner in the first place (as opposed to simply passing it through, which is easier said than done), which is not optimal.

  My solution to this problem was to create a method (subdivider::edges_only) that only constructs edges from the incoming mesh data, and can handle faces containing an arbitrary number of vertices. This method only runs if maxlevel <= 0.

  In terms of Sverchok/Blender, the intention here is to have the node generate valid topology when it is not muted and the subdivision level is 0. When the node is muted, the incoming mesh data should be passed directly through (with no edges created), but this can be implemented on the python side.

  Note that I would use the edges_only function for maxlevel > 0, but the refined face vertices (fverts) are extracted one-by-one (from reflastlevel), so this would be slightly tricky to implement, or at least require a for-loop each for extracting the refined faces and constructing the edges from them, where the current solution does both operations in the same loop. For these reasons, I have stuck with the current approach, even though it lacks elegance.

• Testing is handled a little better now, with the test function integrated directly into the module now. The proper way is still unit tests and CI/CD, but I think this is a step in the right direction. Testing example in this readme have been updated accordingly.

Code

• pyOpenSubdiv @ PyPi
  • pip install pyOpenSubdiv
• Discuss

Uploading to PyPi

I followed this guide. Pretty simple, at any rate:

python3 setup.py sdist bdist_wheel
twine upload dist/*

Note that when updating, it looks like it is necessary to either increment the version number in setup.py, or delete the existing version on PyPi and reupload.

Building

General requirements:

• CMake
• Python 3.8+
• Git

Building on Linux

1. Install General Requirements.

2. Install GLFW (Optional, but makes building OpenSubdiv smoother)

   sudo apt-get install libglfw3 libglwf3-dev
   

   This should create/install the files libglfw3.a at /usr/local/lib/ and gflw3.h at /usr/local/include/GLFW/.

3. Install doxygen and xorg-dev (possibly optional)

   sudo apt-get install doxygen xorg-dev
   

4. Clone OpenSubdiv

   git clone https://github.com/PixarAnimationStudios/OpenSubdiv.git
   

5. Build and Install OpenSubdiv

   $ cd OpenSubdiv 
   $ mkdir build 
   $ cd build
   
   $ cmake -D NO_PTEX=1 -D NO_DOC=1 -D NO_OMP=1 -D NO_TBB=1 -D NO_CUDA=1 -D NO_OPENCL=1 -D NO_CLEW=1 -D GLFW_LOCATION="/usr/" ..
   
   $ cmake --build . --config Release --target install

   This should create an opensubdiv/ directory at /usr/local/include/

   .
   └── opensubdiv/
       ├── far/  
       ├── hbr/   
       ├── osd/   
       ├── sdc/   
       ├── version.h
       └── vtr/
   

   And libosdCPU and libosdGPU library files at /usr/local/lib/

   .
   ├── libosdCPU.a
   ├── libosdCPU.so -> libosdCPU.so.3.4.4
   ├── libosdCPU.so.3.4.4
   ├── libosdGPU.a
   ├── libosdGPU.so -> libosdGPU.so.3.4.4
   └── libosdGPU.so.3.4.4
   

6. Compile

   This does not work (does not compile the osd libraries statically? Throws the error OSError: libosdCPU.so.3.4.4: cannot open shared object file: No such file or directory when trying to import ctypes_OpenSubdiv.so in e.g. load_library.py)

   g++ ctypes_subdivider.cpp -losdGPU -losdCPU -o ctypes_OpenSubdiv.so -fPIC -shared
   

   This seems to work (compiles the static osd .a libraries into ctypes_OpenSubdiv.so?) (compile syntax courtesy this SE answer)

   g++ ctypes_subdivider.cpp -L/usr/local/lib/ -l:libosdGPU.a -l:libosdCPU.a -o ctypes_OpenSubdiv.so -fPIC -shared
   

   Note that -L/usr/local/lib/ is unnecessary, but included for reference purposes.

   Some additional convenience commands:

   Compile directly into pyOpenSubdiv/clib:

   g++ ctypes_subdivider.cpp -L/usr/local/lib/ -l:libosdGPU.a -l:libosdCPU.a -o package/pyOpenSubdiv/clib/ctypes_OpenSubdiv.so -fPIC -shared
   

   Compile to executable:

   g++ ctypes_subdivider.cpp -L/usr/local/lib/ -l:libosdGPU.a -l:libosdCPU.a -o ctypes_OpenSubdiv_executable
   

7. Test:

   Testing pyOpenSubdiv on Linux now uses docker, which I installed following this guide.

   With docker installed, build the test image by running docker build -t ubuntu_docker . from within the dockers directory. Then, start the container by running start_docker.sh from the root of this repo.

   $ ./start_docker.sh 

   The start_docker.sh shell script automatically installs the pyOpenSubdiv package to the container, and on exit deletes miscellaneous directories created from installing and running the module, and the container (note that deleting the module directories requires sudo).

   Start IPython (run $ ipython at the container terminal) (note that using IPython is technically optional, and it is only my preference to use it), import the pyOpenSubdiv module, load pyOpenSubdiv.pysubdivision, and run pysubdivision.test_pysubdivide()

   In [2]: import pyOpenSubdiv
   In [3]: from pyOpenSubdiv import pysubdivision
   In [4]: pysubdivision.test_pysubdivide()
   

   pysubdivision.test_pysubdivide() should produce an output like the following (the Runtime tests will more than likely be different between instances and hardware)

   In [4]: pysubdivision.test_pysubdivide()
   Subdivision Tests
   cube @ 0
   suzanne @ 0
   triangles @ 0
   ngons @ 0
   ngons2 @ 0
   cube @ 1
   suzanne @ 1
   triangles @ 1
   ngons @ 1
   ngons2 @ 1
   cube @ 2
   suzanne @ 2
   triangles @ 2
   ngons @ 2
   ngons2 @ 2
   
   Verbose Test
   maxlevel 1
   New Vertices 26
   v -0.277778 -0.277778 0.277778
   v 0.277778 -0.277778 0.277778
   v -0.277778 0.277778 0.277778
   v 0.277778 0.277778 0.277778
   v -0.277778 0.277778 -0.277778
   v 0.277778 0.277778 -0.277778
   v -0.277778 -0.277778 -0.277778
   v 0.277778 -0.277778 -0.277778
   v 0.000000 0.000000 0.500000
   v 0.000000 0.500000 0.000000
   v 0.000000 0.000000 -0.500000
   v 0.000000 -0.500000 0.000000
   v 0.500000 0.000000 0.000000
   v -0.500000 0.000000 0.000000
   v 0.000000 -0.375000 0.375000
   v 0.375000 0.000000 0.375000
   v 0.000000 0.375000 0.375000
   v -0.375000 0.000000 0.375000
   v 0.375000 0.375000 0.000000
   v 0.000000 0.375000 -0.375000
   v -0.375000 0.375000 0.000000
   v 0.375000 0.000000 -0.375000
   v 0.000000 -0.375000 -0.375000
   v -0.375000 0.000000 -0.375000
   v 0.375000 -0.375000 0.000000
   v -0.375000 -0.375000 0.000000
   e 0 14
   e 0 25
   e 1 15
   e 2 17
   e 3 16
   e 3 18
   e 4 20
   e 5 19
   e 5 21
   e 6 23
   e 7 22
   e 7 24
   e 8 17
   e 9 20
   e 10 23
   e 11 25
   e 12 15
   e 13 23
   e 14 8
   e 14 1
   e 14 11
   e 15 8
   e 15 3
   e 16 8
   e 16 2
   e 16 9
   e 17 0
   e 17 13
   e 18 9
   e 18 5
   e 18 12
   e 19 9
   e 19 4
   e 19 10
   e 20 2
   e 20 13
   e 21 10
   e 21 7
   e 21 12
   e 22 10
   e 22 6
   e 22 11
   e 23 4
   e 24 11
   e 24 1
   e 24 12
   e 25 6
   e 25 13
   f 1 15 9 18
   f 15 2 16 9
   f 9 16 4 17
   f 18 9 17 3
   f 3 17 10 21
   f 17 4 19 10
   f 10 19 6 20
   f 21 10 20 5
   f 5 20 11 24
   f 20 6 22 11
   f 11 22 8 23
   f 24 11 23 7
   f 7 23 12 26
   f 23 8 25 12
   f 12 25 2 15
   f 26 12 15 1
   f 2 25 13 16
   f 25 8 22 13
   f 13 22 6 19
   f 16 13 19 4
   f 7 26 14 24
   f 26 1 18 14
   f 14 18 3 21
   f 24 14 21 5
   
   Runtime
   Cube: 0.478s @ 0 x 10000 
   Cube: 1.941s @ 1 x 10000 
   Cube: 5.102s @ 2 x 10000 
   
   Output
   v [-0.2777777910232544, -0.2777777910232544, 0.2777777910232544]
   v [0.2777777910232544, -0.2777777910232544, 0.2777777910232544]
   v [-0.2777777910232544, 0.2777777910232544, 0.2777777910232544]
   v [0.2777777910232544, 0.2777777910232544, 0.2777777910232544]
   v [-0.2777777910232544, 0.2777777910232544, -0.2777777910232544]
   v [0.2777777910232544, 0.2777777910232544, -0.2777777910232544]
   v [-0.2777777910232544, -0.2777777910232544, -0.2777777910232544]
   v [0.2777777910232544, -0.2777777910232544, -0.2777777910232544]
   v [0.0, 0.0, 0.5]
   v [0.0, 0.5, 0.0]
   ...
   e [0, 14]
   e [0, 25]
   e [1, 15]
   e [2, 17]
   e [3, 16]
   e [3, 18]
   e [4, 20]
   e [5, 19]
   e [5, 21]
   e [6, 23]
   ...
   f [0, 14, 8, 17]
   f [14, 1, 15, 8]
   f [8, 15, 3, 16]
   f [17, 8, 16, 2]
   f [2, 16, 9, 20]
   f [16, 3, 18, 9]
   f [9, 18, 5, 19]
   f [20, 9, 19, 4]
   f [4, 19, 10, 23]
   f [19, 5, 21, 10]
   ...
   
   static float g_verts[8][3] = { {-0.50f, -0.50f, 0.50f},
   {0.50f, -0.50f, 0.50f},
   {-0.50f, 0.50f, 0.50f},
   {0.50f, 0.50f, 0.50f},
   {-0.50f, 0.50f, -0.50f},
   {0.50f, 0.50f, -0.50f},
   {-0.50f, -0.50f, -0.50f},
   {0.50f, -0.50f, -0.50f} };
   
   static int g_nverts = 8, g_nfaces = 6;
   
   static int g_vertsperface[6] = {4, 4, 4, 4, 4, 4};
   
   static int g_vertIndices[24] = {
   0, 1, 3, 2,
   2, 3, 5, 4,
   4, 5, 7, 6,
   6, 7, 1, 0,
   1, 7, 5, 3,
   6, 0, 2, 4 };
   

   The pyOpenSubdiv module may be uninstalled by running the command python -m pip uninstall pyOpenSubdiv.

Building on Windows (Visual Studio)

1. Install General Requirements.

2. Install GLFW (Optional, but makes building OpenSubdiv smoother)

   • Download the GLFW Windows pre-compiled binaries (64-bit), and unzip file.

3. Clone OpenSubdiv

   git clone https://github.com/PixarAnimationStudios/OpenSubdiv.git
   

4. Build and Install OpenSubdiv
   From an administrator powershell (vary appropriately for Visual Studio version and GLFW path):

   $ cd OpenSubdiv 
   $ mkdir build 
   $ cd build
   
   $ cmake ^ -DCMAKE_GENERATOR_PLATFORM=x64 -G "<Visual Studio Version>" ^ -D NO_PTEX=1 -D NO_DOC=1 ^ -D NO_OMP=1 -D NO_TBB=1 -D NO_CUDA=1 -D NO_OPENCL=1 -D NO_CLEW=1 ^ -D "GLFW_LOCATION=<path to GLFW>" ^ ..
   
   $ cmake --build . --config Release --target install

   For example

   # Example build command 
   $ cmake ^ -DCMAKE_GENERATOR_PLATFORM=x64 -G "Visual Studio 17 2022" ^ -D NO_PTEX=1 -D NO_DOC=1 ^ -D NO_OMP=1 -D NO_TBB=1 -D NO_CUDA=1 -D NO_OPENCL=1 -D NO_CLEW=1 ^ -D "GLFW_LOCATION=C:/Users/<username>/Desktop/cpp/glfw-3.3.7.bin.WIN64/glfw-3.3.7.bin.WIN64" ^ ..

   A successful build and install should create an opensubdiv\ directory at C:\Program Files\OpenSubdiv\include\ with the following structure

   C:.
   ├───far
   ├───hbr
   ├───osd
   ├───sdc
   └───vtr
   

   And libosdCPU.lib and libosdGPU.lib library files at C:\Program Files\OpenSubdiv\lib\

5. Install Visual Studio Community, include Desktop Development with C++.

6. Create Visual Studio C++ Project

   • Create a new, blank, C++ project in Visual Studio.
   • Configure to target Release and x64.
   • Add ctypes_subdivider.cpp to the Source Files by right-clicking Source Files in the Solution Explorer, selecting Add->Existing Item..., and navigating to ctypes_subdivider.cpp in the browser window.
   • Note: It is also possible to make a C++ project in this directory (the repository) with a Source.cpp file containing just the line

     #include "../ctypes_subdivider.cpp"

     The following setup steps remain the same.

7. Configure solution properties (for All builds, All platforms)

   1. Set the solution Configuration Type to Dynamic Library (.dll)

      

      Right-click Solution -> Properties -> Configuration Properties -> Configuration Type -> Dynamic Library (.dll)
      

   2. Add OpenSubdiv\include to Additional Include Directories

      

      Right-click Solution -> C/C++ -> General -> Additional Include Directories -> C:\Program Files\OpenSubdiv\include
      

   3. Add Opensubdiv\libs to Additional Library Directories

      

      Right-click Solution -> Properties -> Linker -> General -> Additional Library Directories -> C:\Program Files\OpenSubdiv\lib
      

   4. Add OpenSubdiv library binaries to linker's Additional Dependencies

      

      Right-click Solution -> Properties -> Linker -> Input -> Additional Dependencies -> osdCPU.lib;oscGPU.lib
      

8. Build the solution (Build -> Build Solution), which should create a ctypes_OpenSubdiv.dll file at x64\Release\.

9. Test:

   Testing on Windows does not use a docker container, even though maybe it should, but instead involves directly installing the module from the setup.py file in the package directory. A valid python installation needs to be present on your machine for this test to work.

   Navigate to <this repo>/package, and install pyOpenSubdiv by running the command

   python -m pip install . 

   Then import pyOpenSubdiv, load pyOpenSubdiv.pysubdivision, and run pysubdivision.test_pysubdivide() (again, I am using IPython here, which is optional)

   In [1]: import pyOpenSubdiv
   In [2]: from pyOpenSubdiv import pysubdivision
   In [3]: pysubdivision.test_pysubdivide()
   

   pysubdivision.test_pysubdivide() should produce an output like

   In [3]: pysubdivision.test_pysubdivide()
   Subdivision Tests
   cube @ 0
   suzanne @ 0
   triangles @ 0
   ngons @ 0
   ngons2 @ 0
   cube @ 1
   suzanne @ 1
   triangles @ 1
   ngons @ 1
   ngons2 @ 1
   cube @ 2
   suzanne @ 2
   triangles @ 2
   ngons @ 2
   ngons2 @ 2
   
   Verbose Test
   maxlevel 1
   New Vertices 26
   v -0.277778 -0.277778 0.277778
   v 0.277778 -0.277778 0.277778
   v -0.277778 0.277778 0.277778
   v 0.277778 0.277778 0.277778
   v -0.277778 0.277778 -0.277778
   v 0.277778 0.277778 -0.277778
   v -0.277778 -0.277778 -0.277778
   v 0.277778 -0.277778 -0.277778
   v 0.000000 0.000000 0.500000
   v 0.000000 0.500000 0.000000
   v 0.000000 0.000000 -0.500000
   v 0.000000 -0.500000 0.000000
   v 0.500000 0.000000 0.000000
   v -0.500000 0.000000 0.000000
   v 0.000000 -0.375000 0.375000
   v 0.375000 0.000000 0.375000
   v 0.000000 0.375000 0.375000
   v -0.375000 0.000000 0.375000
   v 0.375000 0.375000 0.000000
   v 0.000000 0.375000 -0.375000
   v -0.375000 0.375000 0.000000
   v 0.375000 0.000000 -0.375000
   v 0.000000 -0.375000 -0.375000
   v -0.375000 0.000000 -0.375000
   v 0.375000 -0.375000 0.000000
   v -0.375000 -0.375000 0.000000
   e 0 14
   e 0 25
   e 1 15
   e 2 17
   e 3 16
   e 3 18
   e 4 20
   e 5 19
   e 5 21
   e 6 23
   e 7 22
   e 7 24
   e 8 17
   e 9 20
   e 10 23
   e 11 25
   e 12 15
   e 13 23
   e 14 8
   e 14 1
   e 14 11
   e 15 8
   e 15 3
   e 16 8
   e 16 2
   e 16 9
   e 17 0
   e 17 13
   e 18 9
   e 18 5
   e 18 12
   e 19 9
   e 19 4
   e 19 10
   e 20 2
   e 20 13
   e 21 10
   e 21 7
   e 21 12
   e 22 10
   e 22 6
   e 22 11
   e 23 4
   e 24 11
   e 24 1
   e 24 12
   e 25 6
   e 25 13
   f 1 15 9 18
   f 15 2 16 9
   f 9 16 4 17
   f 18 9 17 3
   f 3 17 10 21
   f 17 4 19 10
   f 10 19 6 20
   f 21 10 20 5
   f 5 20 11 24
   f 20 6 22 11
   f 11 22 8 23
   f 24 11 23 7
   f 7 23 12 26
   f 23 8 25 12
   f 12 25 2 15
   f 26 12 15 1
   f 2 25 13 16
   f 25 8 22 13
   f 13 22 6 19
   f 16 13 19 4
   f 7 26 14 24
   f 26 1 18 14
   f 14 18 3 21
   f 24 14 21 5
   
   Runtime
   Cube: 0.104s @ 0 x 10000
   Cube: 0.272s @ 1 x 10000
   Cube: 0.620s @ 2 x 10000
   
   Output
   v [-0.2777777910232544, -0.2777777910232544, 0.2777777910232544]
   v [0.2777777910232544, -0.2777777910232544, 0.2777777910232544]
   v [-0.2777777910232544, 0.2777777910232544, 0.2777777910232544]
   v [0.2777777910232544, 0.2777777910232544, 0.2777777910232544]
   v [-0.2777777910232544, 0.2777777910232544, -0.2777777910232544]
   v [0.2777777910232544, 0.2777777910232544, -0.2777777910232544]
   v [-0.2777777910232544, -0.2777777910232544, -0.2777777910232544]
   v [0.2777777910232544, -0.2777777910232544, -0.2777777910232544]
   v [0.0, 0.0, 0.5]
   v [0.0, 0.5, 0.0]
   ...
   e [0, 14]
   e [0, 25]
   e [1, 15]
   e [2, 17]
   e [3, 16]
   e [3, 18]
   e [4, 20]
   e [5, 19]
   e [5, 21]
   e [6, 23]
   ...
   f [0, 14, 8, 17]
   f [14, 1, 15, 8]
   f [8, 15, 3, 16]
   f [17, 8, 16, 2]
   f [2, 16, 9, 20]
   f [16, 3, 18, 9]
   f [9, 18, 5, 19]
   f [20, 9, 19, 4]
   f [4, 19, 10, 23]
   f [19, 5, 21, 10]
   ...
   
   static float g_verts[8][3] = { {-0.50f, -0.50f, 0.50f},
   {0.50f, -0.50f, 0.50f},
   {-0.50f, 0.50f, 0.50f},
   {0.50f, 0.50f, 0.50f},
   {-0.50f, 0.50f, -0.50f},
   {0.50f, 0.50f, -0.50f},
   {-0.50f, -0.50f, -0.50f},
   {0.50f, -0.50f, -0.50f} };
   
   static int g_nverts = 8, g_nfaces = 6;
   
   static int g_vertsperface[6] = {4, 4, 4, 4, 4, 4};
   
   static int g_vertIndices[24] = {
   0, 1, 3, 2,
   2, 3, 5, 4,
   4, 5, 7, 6,
   6, 7, 1, 0,
   1, 7, 5, 3,
   6, 0, 2, 4 };
   

Sverchok Integration

• These links are useful for dealing with syncing forks up with master repos.
  • How do I update or sync a forked repository on GitHub?
  • Git Merge Master into Branch
  • Also the .git/config file needs to be configured properly. It's all very confusing and annoying and I wish it made more sense.
  • I think the sync process is like this

git fetch upstream 
git checkout master 
git rebase upstream/master 
git checkout <branch> (e.g. sverchok_OpenSubdiv_2)
git merge <branch to merge> (e.g. master)
git push

• Make sure that pyOpenSubdiv is not installed for Blender's python

  • Blender's python is at C:\Program Files (x86)\Steam\steamapps\common\Blender\3.4\python\bin> (or similar)
  • Installed packages are at C:\Program Files (x86)\Steam\steamapps\common\Blender\3.4\python\lib\site-packages
  • pyOpenSubdiv may be uninstalled by running e.g. ./python.exe -m pip uninstall pyOpenSubdiv from within the C:\Program Files (x86)\Steam\steamapps\common\Blender\3.4\python\bin> (or similar) folder.

• Rename: Catmull-Clark Subdivision

• Change bl_idname back to "SvOpenSubdivisionNode"

• nodes/modifier_change/opensubdivision.py

  • Changing the name to Catmull-Clark Subdivision, but I'm leaving the python script as opensubdivision.py.
  • The Catmull-Clark Subdivision node implementation.
  • Imports pyOpenSubdiv.
  • SvCatmullClarkSubdivisionNode -> SvOpenSubdivisionNode
  • opensubdivision.py

• (No Change) dependencies.py

  • Add pyOpenSubdiv as Sverchok optional dependency.

    # dependencies.py
    ...
    pyOpenSubdiv_d = sv_dependencies["pyOpenSubdiv"] = SvDependency("pyOpenSubdiv","https://github.com/GeneralPancakeMSTR/pyOpenSubdivision")
    pyOpenSubdiv_d.pip_installable = True
    try:
        import pyOpenSubdiv
        pyOpenSubdiv_d.message = "pyOpenSubdiv package is available"
        pyOpenSubdiv_d.module = pyOpenSubdiv
    except ImportError:
        pyOpenSubdiv_d.message = "sv: pyOpenSubdiv package is not available, the Catmull-Clark Subdivision node will not be available"
        info(pyOpenSubdiv_d.message)
        pyOpenSubdiv = None 
    ...

• (No Change) settings.py

  • Draw the pyOpenSubdiv dependency installation box in the Sverchok Preferences "Extra Nodes" tab.

    # settings.py
    ...
    draw_message(box, "scipy")
    draw_message(box, "geomdl")
    draw_message(box, "skimage")
    draw_message(box, "mcubes")
    draw_message(box, "circlify")
    draw_message(box, "cython")
    draw_message(box, "numba")
    draw_message(box, "pyOpenSubdiv") # Add option to install pyOpenSubdiv
    ...

• index.md

  • Add the Catmull-Clark Subdivision node to the Sverchok node menu.

    ## Modifier Make
        LineConnectNodeMK2
        ---
        SvCatmullClarkSubdivisionNode -> SvOpenSubdivisionNode
        SvSubdivideNodeMK2
        SvSubdivideToQuadsNode
        SvOffsetLineNode
        SvContourNode
        ---
        SvDualMeshNode
        SvDiamondMeshNode
        SvClipVertsNode
        ---
        SvBevelCurveNode
        SvAdaptiveEdgeNode
        SvAdaptivePolygonsNodeMk3
        SvDuplicateAlongEdgeNode
        SvFractalCurveNode
        SvFrameworkNode
        SvSolidifyNodeMk2
        SvWireframeNode
        SvPipeNode
        SvMatrixTubeNode

• docs/nodes/modifier_change/modifier_change_index.rst

  • Added opensubdivision node.

• docs/nodes/modifier_change/opensubdivision.rst

  • Catmull-Clark Subdivision (opensubdivision.py) node documentation.
  • Install successfully (old node)
    
  • Test Node (old node)
    
  • New tests, with renamed ("Catmull-Clark Subdivision") node, demonstrating better case handling.
    
    
    
    

ToDo

Building on OS X

• Implement

Error Handling and Crash Prevention

• Subdivision <= 0 should be handled better now.
• Identify and resolve new bugs.
• Handle case where user accidentally inputs a Face list that refers to vertices that do not exist in the vertices list (e.g. Faces refer to vert 7, where there are only 5 vertices).

CI/CD Build and Deploy pipeline

• Implement

Unit Testing

• Implement