When to start with a walking skeleton

This is an unpublished post. It's likely (more) full of spelling mistakes and errors. It may also end abruptly without a conclusion.

TDLR:

If I don’t even know that I’ll need a web service/whatever I don’t worry about a walking skeleton. I hack together anything that takes input and gives output and helps me explore a concept. That way I don’t waste time with infrastructure that I might not need.

If I know I’m definitely going to be deploying an API of some kind then I start with getting a walking skeleton for that kind of API up and running. That way I know I’ll always be able to deploy my work.

Thoughts? Comments? Send me a tweet!