I have tried something completely against to current web trends.
Edit: I don't recommend implement this method method unless you're implementing an undo-redo for something it's not possible to make a deep copy after each action. Initially I published this as an NPM package and published to code on github but I deleted it. I think it's better if you go for a case driven according your application and without using any framework or a library, by taking all the state transitions of your application into the account, and implement.
As you may already know, HTTP is a stateless protocol. The communication between the client and server happens with some flags sent by client and server mutually. It means, we actually don't really know which user is authenticated. User is sending their session id via HTTP cookies and we (server) retrieve the related record from the database, and prepare the page response with the information of authenticated user.
I have been trying serverless computing platforms in last years. Basically these platforms allows you to publish functions and execute them periodically, or by triggering it with an HTTP request.
In my previous blog-post, I created a simple http server that calls a redis method; increases the specified key with request body in the redis store.
AsyncIO is a built-in python library to write concurrent code with a new language keywords "async" and "await'. AsyncIO landed into python in 3.3 version. Since the Python2 is no longer supported, AsyncIO will be our swiss army knife to handle