Composability is the main one for me that's interesting. When serving data, I often use HTTP pulling from a Postgres DB. Both of those are massively over engin

Composability is the main one for me that's interesting. When serving data, I often use HTTP pulling from a Postgres DB. Both of those are massively over engineered for the problems I want to solve but I use them because: * They're already built so it's not my engineering time. * Others can interact with them easily. This is more strongly the case, and becoming more so over time, in crypto work. I can make something with key based auth that I can uniquely identify such that a user can prove they own it to integrate in my "real world" systems, while letting them sell it, trade it, auction it or whatever. To build this I do a starter tutorial, load a common library and publish. After I can just use already existing APIs for integration. That's it and it works because the data and APIs and Auth are all common and standardised now (either inherently for Auth or by convention for the data).