~A Jupyter environment for microsystem design and multiphysics simulation~

This is a rewrite of the author's prior attempts at making devices entirely in Python. The desired goal is a usable toolset capable of design, modeling, and FEA tasks using only open-source components.

• An installation of Podman or Docker capable of running Linux containers.

1. Build the container with Podman: podman build --format docker -t vivicad-ng <repo dir> Or if using Docker: docker build -t vivicad-ng <repo dir>

2. cd to the desired run location, such as another project. Run the container, specifying a port and a working directory. Here, port 8888 and the current directory are selected: podman run -d -p 8888:8888 -v ${PWD}:/mnt -t vivicad-ng

3. Access the container's JupyterLab server: xdg-open http://localhost:8888 (Or just open the URL in your browser!) In Windows, replace xdg-open with start. In OSX, replace it with open.

4. Note that authentication for Jupyter is disabled for developer convenience. Do not expose the server to the Internet!!

Q: How do I kill a running container?

A: List and find the name of the container with podman ps. Then, run podman kill <container name>.

Q: I am getting errors regarding disk space. How do I make more space?

A: Delete old containers. To delete everything: podman rm -a && podman rmi -a

Q: How do I run other notebooks with this container?

A: Change directory to the external notebook location then start the container. ${PWD}:/mnt in running the container takes the current directory and mounts it in the container at /mnt. Alternatively, ${PWD} can be replaced with an absolute path or appended with a subdirectory, e.g. ${PWD}/notebooks.

Q: How do I fix Permission Denied errors when accessing files in a mounted directory?

A: Run sudo chmod 777 <directory path> on the mounted directory to allow the container to modify the contents.

Q: How do I add dependencies to this project?

A: There are two places to add dependencies. Native programs, dependencies NOT installable by pip / Poetry, or anything not used by the vivicad codebase should be installed by Micromamba and thus written to the environment.yaml file. Dependencies in use by the codebase and installable by pip / Poetry should be added to pyproject.toml file with poetry add <package name>.

Q: How is this software structured?

A: This software consists of a Python package and a container providing a JupyterLab interface to the package and third-party software meant to be used along with the package. The package is for writing an API to be used within Jupyter or by other software. The container provides a portable and convenient environment for an end-user to interact with the package and included third-party software.

Q: How is the container built?

A: The container is defined by the Containerfile file. This file is configured to install necessary runtime software within the container. Included software is managed by Micromamba, a Conda-compatible package manager, and the container is built on top of micromamba-docker. Conda packages are declared in environment.yaml. As a last step, the vivicad-ng Python package is installed within the package. Running the package will start a JupyterLab server at port 8888.