Should it all really just live in one map? Should we have protocol methods that get certain important things so implementors have a clearer sense of what they need to do? Or just document what things should be there and what shouldn't? Has implications for #3, as well as the simulator implementation.
Should it all really just live in one map? Should we have protocol methods that get certain important things so implementors have a clearer sense of what they need to do? Or just document what things should be there and what shouldn't? Has implications for #3, as well as the simulator implementation.