Communication and organization. Tricky in real-life and tricky in code; keeping all the data flowing between layers of the application, through APIs, from data stores, through analyzers, with notifications, etc, etc.
Raw experience will probably help a lot here, but I find myself going back and forth on naming conventions. Proper naming should describe what you want to do, and from that, what kind of data you would need. There are classes that I've refactored about ten times now, just to refine its place and its data.
Putting all those classes in the right context helps a lot in defining the right data needed and putting all that data on the right communications bus. It's easy to overdo and send way too much information around, but it's also very, very irritating if you keep finding different missing pieces along the way. Working towards a balanced data set just means to keep shaving and keep refactoring. Just like with those pesky class names...
Done
- Structured status reporting 📄
Upcoming
- More pipelines 🚰
