There definitely seems to be a sweet spot where projects are challenging enough to be engaging yet familiar enough that you don't get frustrated. Finding that window looks to be key.

It's also a great idea to put a deadline for yourself, say, having a small MVP working within a month, and then that's that. Having something 'accomplished' in that timeframe could be neat, as in, it would likely spur me to keep developing it once I see it working. Thank you.