DEA Tools is intended to be an accessible open source package, and many of the tools there are theoretically useful outside of DEA. However, a big issue is we don't really know easy it is to install by external users outside the Sandbox. This is largely due to inexperience with Python packaging across DEA Notebooks/Tools developers (our current setup has barely changed since it was originally set up several years ago). Recently we've received feedback from USGS that they tried to use the package but were thwarted by a host of dependency issues. This isn't a great look, especially if we want to spruik new DEA Tools content more widely with a new version.
We should review our current DEA Tools Python packaging setup, to ensure the package is as easy to install as possible by external users outside the Sandbox. The end goal is that an external user can easily do pip install dea-tools, and have the package installed as smoothly as possible.
DEA Tools is intended to be an accessible open source package, and many of the tools there are theoretically useful outside of DEA. However, a big issue is we don't really know easy it is to install by external users outside the Sandbox. This is largely due to inexperience with Python packaging across DEA Notebooks/Tools developers (our current setup has barely changed since it was originally set up several years ago). Recently we've received feedback from USGS that they tried to use the package but were thwarted by a host of dependency issues. This isn't a great look, especially if we want to spruik new DEA Tools content more widely with a new version.
We should review our current DEA Tools Python packaging setup, to ensure the package is as easy to install as possible by external users outside the Sandbox. The end goal is that an external user can easily do
pip install dea-tools, and have the package installed as smoothly as possible.