It's an interesting question I've been pondering for several years now. For me, it's solving known problems (e.g., a web app, a data investigation app, etc.) an

It's an interesting question I've been pondering for several years now. For me, it's solving known problems (e.g., a web app, a data investigation app, etc.) and evolving the code based on best practices and what the language/framework/etc. allow. It's a more complicated question than, say, learning a musical instrument, because there aren't hard-and-fast benchmarks beyond solving well-known itches. But I think solving those well-known itches goes a long way to formalizing what it means to deliberately practice development.