My playground for testing and or learning various libraries and code techniques.
I made this repo with the idea of having a shared contract between client and server and testing out a variety of languages and frameworks.
I got the idea when I encountered hono which I had not used before and was already playing with a rust server so thought this would be a good way to proceed.
Nostalgia mostly but also its a free open API that has a good data structure and a good amount of helper libraries. This allows me to code in a way I would for production code. I am not writing a tool to contact AWS directly, instead I would use the provided toolkit for the language. Pokémon is also completely non work related and light-hearted, this means no one will confuse it for prod code when it is not meant to be.
A quick note about the clients: they do not share a UX, I will undoubtedly port some components or snippets of code across to re-use and save time but some will be too different to easily recreate. This will mean that I wont be able to have one suite of e2e tests that can run on any client but it is something that I intend to test when I have everything up to date with the current contract.
This repo is about testing new technologies and that includes AI. I am largely not that interested in the raw html and css except where is is substantially different from what I am used to. This means that I will probably use some AI tools to automagically generate the UX in some places. I will also be using AI to generate tests, as a testing specialist I am interested to see how well AI can handle that task, from what I have seen so far its not good but I will give it a fair run and try in a few of the sub-projects.
I am starting with MongoDB but will potentially add other storage systems in the future.
I could do it all in memory but in the docs of Pokeapi they do request that people cache locally when possible to prevent overtaxing of their servers, fair enough.
The plan is that any of the servers can start up and write and pull data from the same MongoDB store which will for now just be run in docker.
The contract between client and server has been pretty much nailed down now.
Are all of the clients and servers up to date? ...no. Will they be? Yes, absolutely, I want this to be a living repo that will almost always build and run with any combination of client and server.
I have preferences and biases that's for sure, I really enjoy working with the react-client and rust-server more than others so they will most likely be the most up to date ones.
Short answer yes. The long answer is that I would like to get to the point where I can load all of the clients and servers in docker compose and have a switch of some sort to select which server to use from the clients.
Eventually I would like to have a similar setup using something like minikube to help me get better at kubernetes and another setup so it can all be run from a nix flake. When I eventually get a machine to use as a server I will experiment with automatic deployments, probably using nix at first.
I am toying with the idea of the following clients:
Electronwrapped around eitherreact-clientorremix-client. And the following servers:- Java
- C#
- Python
If you have read this far and have any suggestions for frameworks and technologies I should try raise an issue and ill take a look. If you like the idea of this repo and want to do one yourself go for it, I have found it extremely useful. Feel free to also fork this repo and push back any tech you want to add. I intend to have the shape of the client / server contract on this readme when it is a bit more mature.
- Pencil client
- Pyhton server
- Setup proper linting for each project
- Finish go-server
- Hapi server
- TypeORM
- RxJS
- Finish svelte-client
- update all to use current contract
- document current contract here in this readme
- add e2e tests for some of the clients
- start looking into infrastructure