In the list of micro-frameworks, please add Nameko - it's a microservices framework with best practices applied for software testing.

Please add web2py , it's a wonderful framework. Especially the form creation and handling, pydal for non-intrusive backend-independent database abstraction.

Not that I use it (too used to relational models) but probably it's useful in a year or two when the first bugs are out

for future reference: sqlite. serverless sql databases offering in-process speed and low overhead. Especially good for loads of reads (cached stats, scores, historical data etc), less interesting for heavy write-to databases (session databases, chat storage) due to lack of proper locking required for multi-write access. But quite good for development and quad solutions.

Can you elaborate a bit on what you are trying to achieve? isn't from lib import x,y,z what you are searching for? If it's dynamic loading you need, you shouldn't use __import__ but use importlib.import_module ( link ) instead

I'm using Graphene (the python port) and stuck it behind Flask (web application server) which hides behind gunicorn (WSGI server, a python generic webapplication interface) which in turn is behind api-umbrella (an api reverse proxy). My main issue at the moment is appropriate caching when considering invalidation and retrieving different related details from different microservices. The dataloader is great but also a headache generator in that sense. But this is more backend related than graphql related. More graphql related would be how to do proper versioning (remaining backward compatibility) and have different customers (api consumers/frontends) use the same api with slightly different requirements. You could say require different schema's, but that might not be entirely true. Still working on clarity on that. Big pains for me was the lack of proper documentation in graphene. Also the use of Promises was new to me, and seems less natural in python than in javascript. Since I stick with python there is only graphene and no other library to choose from that compares. (unless someone can point me a library I missed)

Heavy Evernote paid subscription user here. Why paid? because it indexes images, pdfs, images inside pdf's etc. It has easy integration with a lot of other tools and services, doesn't have a Linux client (for which i blame them) but still i consider it my best option so far. One picture of a bookshelf and i have indexed all the titles. (yes, i too use books) Whiteboard and mobile taken images of paper are enhanced and readable. I mostly use Camscanner for the image processing on Android (better quality, good pdf export and batch support) I index quite a few blog posts, forum answers, Hashnode items, because i have seen too many good ones gone to waste with just a bookmark. plus, bookmarks don't index content and i remember part of the content, not the title. references to moving parts (which i don't want to index) still become bookmarks in chrome but many times end up both in chrome and in Evernote. Because chrome allows me to search in the address bar, which still is quicker than switching to another window. As much as i like evernote over all the other solutions i found so far, it's still far from perfect. Especially the UI and lack of complex search can be a pain. Sometimes it feels like an interconnected and smart notepad++ that simply refuses to take the leap... But hey, combined with the other stuff and the API's nowadays, we just hack our own solutions right? Complete projects go into dropbox because of the file sync obviously and the use of jetbrains tools for code. (pycharm with markdown support and "live" PlantUML is really nice) A private Rhodecode server holds a lot of code in both git and mercurial. I check out my repo's in a dropbox (linked) folder so i can continue where i left of, and have an automatic backup. Project related documentation is in the sourcecode, or on a self-hosted Taiga server. Communication about that, and some snippets move through Flock. And my latest weapon of mass creation is a Wacom Bamboo Slate. Because i still ended up with a lot of paper notebooks everywhere. Mostly because hand written notes and schema's are just so darn quick compared to everything digital. But i ended up with 20 notebooks and the management of those physical things was getting tedious. Also, that stack of paper notebooks is quite heavy. So i'm trying the slate now, which holds a paper notebook but automagically produces an electronic version. Of course it's synced with my phone and from there i can export to evernote. It has tags support, like writing #tag on paper, so i'm working to see if i can get those to evernote. Hope this helps.

I remember some of those, see https://www.nexthink.com/blog/natural-language-interfaces-to-databases-nlidb/ for some history about the topic.

Api-umbrella , open source and does the job for me. Nicely configurable, only thing for me is that it doesn't run on arm hardware. Uses NGINX and others. I use it for a graphql backend and some reverse proxying to specific webhook microservices.

Can't you start working on changing the backend first? Also, before you dive completely into a REST API, have you considered Graphql? Though it adds some complexity i think it's easier to maintain than standard restful interfaces, and might require a lot less bandwidth, especially over time.