Stack overflow questions about topic helps a lot.
Examples and small iterative experiments about topic.
Doing lot of groundwork or due-diligence of what you are about to learn before learning is a crucial step. Specially if you are a solo/small-team. Because the cost to learn something which eventually will not fit/work/be-easy-to-learn/has-better-alternatives is costly than learning that itself.
I rely on making diagrams of new concepts : architectural / timing / state-machines / matrix of concepts / quadrants / hierarchy / just enumeration of concepts with one line descriptions against it.