Snailgate

Installation

Clone the repository in the desired location:

git clone https://github.com/caspar/snailgate.git

Go inside the repository:

cd snailgate

Create a virtual enviornment with Python 3.9:

python -m venv venv

Activate the virtual environment:

source venv/bin/activate

Install dependencies:

pip install -r requirements.txt

Build client:

./build_client.sh

Start application (ignore eventlet errors, they do not affect functionality):

./start_server.sh

Access application in any browser at (127.0.0.1 or 0.0.0.0 or localhost at port 5000):

http://127.0.0.1:5000/

Results can be found in the two generated files:

• results_forces.json contains the raw forces across all vertices and all timestamps
• results_data.txt contains the digested results such as the maximum force observed in the system, the timestamp that it happened and the index (which is the corresponding vertex associated with the maximum force)

Build and run local image in Docker

Make sure you have Docker downloaded and have the daemon running. https://docs.docker.com/get-docker/

Build local image:

docker compose build

Run the service in detached mode:

docker compose up -d -V

For codebase updates:

docker compose up -d -V --force-recreate

Pull Image from Docker Hub

Pull client and server images

docker pull xyz010/snailgate-client
docker pull xyz010/snailgate-server

Run the images

docker run xyz010/snailgate-client
docker run xyz010/snailgate-server

UI Stuff

• libigl
• matplotlib
• html canvas

Physics

Input

• Collection of vertices V = (#V, 2) double matrix
• Collection of edges E = (#E, 2) int matrix
• Tags for vertices VP = (#V, 1) int vector(0 = not fixed, 1 = fixed)
• Tags for edges EP = (#E, 1) int vector (0 = spring, 1 = rods, 2 = ropes)
• Level of water double z

Output

• Forces on the nodes F = (#V, 2) double matrix

Future

• Explicit integration (forward Euler w/ Newton's law)
• Quasi-static solution (minimize potential energy of the system). Formulate energy wrt node positions
• Handling of rods: from 2 points coord (4 DOFs) -> 1 point + rotation (3 DOFs)

Recommendations

Separate GUI part from the simulation part.

Response from server

Set of frames with updated positions of verteces

Development

Setup

Install node.js for the javascript build system (webpack).

Install dependencies:

pip install -r requirements.txt

cd ui/client && npm i && cd ../..

Install linting tools globally, so that they work in the development environment:

npm install -g eslint@2.12.0 eslint-plugin-react@5.1.1

Running the application

Start the server:

./start.sh

Open the url http://localhost:5000 in the browser.

Running tests/linting tools

To run the linting tools:

./lint.sh

Deployment

Build the javascript:

./build_client.sh

Start the server in production mode:

./start_server_prod.sh