sandialabs / sibl Goto Github PK

Semantics: actually not the model keys, it is the model values, for collections of models in a single view. So 'model_keys' is semantically wrong. Also, more natural to call out a collection (plural), so use 'models'.

Currently, we plot the edges of a quadrilateral. Joining edges of quads have their edges painted twice during postprocessing. Better to have a point and edge representation of the mesh, and paint each edge once and only once. CBH to implement.

See https://medium.com/opsops/how-to-test-if-name-main-1928367290cb

curve fitting:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5358887/
https://www.geometrictools.com/Documentation/BSplineCurveLeastSquaresFit.pdf
https://www.geometrictools.com/Samples/Mathematics.html
goblet:
https://demonstrations.wolfram.com/PottersWheel/
chess pawn:
http://docs.autodesk.com/3DSMAX/12/ENU/3ds%20Max%202010%20Tutorials/files/WSf742dab041063133-62b9306f112a19e40dd-7ffb.htm
https://www.clear.rice.edu/comp162/assignment02/assignment02.html

Currently dense boundary representations are required because projection snaps the discrete boundary points. We should parameterize the boundary and snap to the closest point on the parameterized boundary, not the discrete boundary points. ACS to implement.

See https://xmldiff.readthedocs.io/ and https://github.com/jrsmith3/diffsvg

To recreate:

1. Set up conda in fish via eval $HOME/miniconda3/bin/conda "shell.fish" "hook" $argv | source
2. Run sibl/config/env_setup.sh

Error:

CommandNotFoundError: Your shell has not been properly configured to use 'conda activate'.
To initialize your shell, run

    $ conda init <SHELL_NAME>

Better specificity for client.

https://docs.python.org/3/library/pathlib.html