The project can be either installed via the included Poetry project configuration or via the PyPI package which can be found [here](https://pypi.org/project/duffy/). 

The package can be easily installed by executing the following command in an activated virtual environment.

```

pip install duffy

```

Be sure to append the command with a version number if you are looking for setting up the project at its state on a specific tag or release. For example, the following command would help specifying the version `3.0.0` for installation.

```

pip install duffy==3.0.0

```

Do note that the pre-release versions of the project are usually not installed implicitly and in order to install them, the version number need to explicitly specified in the manner stated above. This is not true if the pre-release version is the only version available.

```

pip install duffy=3.0.0a0

```

Installing the PyPI package is greatly recommended as it allows you to conveniently upgrade to the newer version of the project without having to keep up with the updates made to the repository. 

As of now, the PyPI package has the following four maintainers.

- [Akashdeep Dhar](https://pypi.org/user/t0xic0der/)

- [Fabian Arrotin](https://pypi.org/user/arrfab/)

- [Nils Philippsen](https://pypi.org/user/nilsph/)

- [Vipul Siddharth](https://pypi.org/user/siddharthvipul/)

The aforesaid maintainers can publish new versions of the project by first, creating a tag of a distinct version number following the semantic versioning scheme and then, using the following command to build and publish packages in succession. Please make sure that you are at the root directory of the project where the `pyproject.toml` and that the project configuration is valid.

```

poetry publish --build

```

Please do note that the mentioned command requires authentication with the use of username and password so as to ensure that only the authorized maintainers can push new releases. It is also greatly recommended to first push a finished release with an alpha tag (say, a finished release `3.0.1` can be pushed to PyPI as a pre-release package `3.0.1a0`) to allow for testing before pushing the release again with the actual version number to not use up the version number if the release requires revisions.