Currently I'm trying to assist an open source project and I suggested that since we are spending lots of time trouble shooting configuration settings over multiple platforms, versions, etc that we consider using a Docker container with our stack preconfigured, so the new contributor (I believe) only needs to spin up that container, clone the repo and start coding.
As someone who has only read about Docker until now; laptop wouldn't support it until I got my new one this week - woohoo - and since it's something I'm excited to try I suggested it. It was accepted as a possible way forward and now I'm stuck with my biggest fear; did I pitch something that isn't a good solution.
... then do we stick with package.json to run the full set up or is there an alternative? I'm fine with sticking it out with package.json, but I was just hoping to simplify. It's important to understand this OSS project is supported mostly by newer devs, so if it's not simplifying for them then it's a no go for us.
It needs to be both the right tool or improvement for the job and make on-boarding of new contributors more *simplistic than now using package.json
What are your thoughts?