Guidelines for adding new Python packages

[raoulcollenteur]

I have restructured how new packages are added. The new format is to add a dictionary with the information about the project to list.json and open a pull request. All the rest will be updated automatically:

    "Pastas": {
        "description": "Analysis of hydrological time series using time series models.",
        "url": "https://github.com/pastas/pastas",
        "category": "Time Series (Analysis)",
        "docs": "https://pastas.readthedocs.io",
        "CI": 1,
        "last_update": "Will be update automatically from conda/pypi",
        "pypi": "https://pypi.org/project/Pastas/",
        "doi_software": "https://doi.org/10.5281/zenodo.1465866",
        "doi_paper": "http://doi.org/10.5281/zenodo.1465866",
        "conda": "https://anaconda.org/conda-forge/Pastas",
        "version": "Will be update automatically from conda/pypi"
    },

I hope this makes everything easier and improves the quality of the list :)

[raoulcollenteur]

Good packages should:

• have documentation (i.e., ReadTheDocs)
• have continuous integration testing (i.e., GH Actions, Travis)
• be deposited in a FAIR repository (i.e., Zenodo)
• be installable with a common Python package manager (i.e., pip, conda, or mamba)
• have an open-source licenses and code repository

In addition:

• Nice-to-have is a software description paper.
• Packages with no releases for the past 5 years are moved to the legacy part.