The sequel (re-write) of Mockingjay-server
- Go version 1.17+
- Docker
make
make run
- I am a much better engineer compared to when I wrote M1J, back in 2015. It was my first Go project. I think I should be able to get more features but have a much simpler codebase than what MJ ended up being.
- A lot of people (relatively speaking) saw value in MJ1, and enjoyed using it, but YAML as a config choice was pure pain. Hoping to do much better here
- Address some of the shortfalls, in particular the need to have some sense of variance between how the local server works, and running the CDC against the real one. Stuff like authentication, API keys, yada yada
- Take a very strict, ATDD approach as I described in learn go with tests
- Leverage example mapping to give myself focus https://github.com/quii/mockingjay-server-two/wiki/Example-mapping
- Build a fast, simple UI where people can dynamically define their server and persist the configuration as they write it. Using https://htmx.org
- MJ1 was super fast, not just at spinning up, but leveraged concurrency to do the CDC checks, which was really nice
- The error messaging in MJ1 wasn't amazing at times, especially when the CDC failed, hope to improve this.
- MJ1 was simple becauuse it was very basic and strict, this was good but at times it meant it was unsuitable for some use-cases. Without turning it into a beast like Pact, I'd like to find a way of letting people add some flexibility to the way they define their contracts
- I think stuff that charmedbracelet have written can help MJ have a friendlier UX on the command-line
- Configuration was the real pain on MJ1. Will look in to using cuelang to see if it can help with this