Hi all, I wanted to share a new pure python implementation of the configurable-http-proxy that I’m hoping will simplify jupyterhub installations for many people (atleast in small environments)
I’ve been using jupyterhub for the past 8+ years, and have found it to be tremendously helpful.
One of the things that has always bugged me is the dependency on nodejs - mainly because most of the servers I manage are python. And I have to additionally install nodejs and configurable-http-proxy just for jupyterhub. While conda/docker ease this - it’s still an overhead, especially for simple installs
So, I started a thread on that: Pure python alternative to configurable-http-proxy · Issue #3528 · jupyterhub/jupyterhub · GitHub
And ended up building my own python implementation of the configurable-http-proxy: configurable-http-proxy · PyPI
The proxy tries to keep the dependencies to a minimal - for example, uses tornado for most of the work.
The main aim being that a simple pip install jupyterhub configurable-http-proxy should get the setup working without any other major dependencies