The most important part of Linus' project was day -1, when he sliced off all the chunks of work that depend on solving open research problems.

You don't want to start your Saturday morning declaring an _action struct, then filling the rest of the day staring at research papers about the current state of fast homomorphic encryption.