When building web3 products, we tend to rely heavily on public testnets, or even building directly on mainnet. often times we also rely on 3rd party services (e.g. subgraphs) which are themselves restricted to live networks, and have their own shortcomings for the development process.

This creates a lot of inefficiencies: difficulty in getting test funds, long feedback loops due to needs to re-deploy all contracts at every step it also makes the communication between contracts and frontend hard to manage efficiently.

In this talk I'll walk through these pitfalls, discuss how we got here, and propose better approaches to achieve a fully reproducible local development setup, and vastly larger feedback loops