open_questions.txt

so, fixing initial map loads of dynamic entities... my options are:

- add (a) new step(s) that trigger after the stale entities check
    - will work, easy enough to make work
    - may have complications
    - won't solve future instances of similar problems
    - hacky; "feels wrong"
- modify the location of the existing init so it happens after stale ent
    - may have complications
    - not straight-forward
- modify the stale ent so it happens before init
    - same as above
- modify the stale ent call so it doesn't clear some stuff
    - hacky
- modify the init call so its stuff won't get cleared by stale ent
    - hacky
- modify the whole system so that stale and init don't clash
    - no idea how to do this
    - a couple days later: did this!
      - got rid of "stale", it was an umbrella to make it so I wouldn't
        have to think about when to change maps and how to load entities,
        at a time when bidirectional communication between managers was
        very hard to do without opening up holes big enough to drive
        dump-trucks through.

        Now that it's easy to (in a limited space) tap a manager on the
        shoulder and ask it nicely to set up the entities, it's easy to
        do so within the change_maps function. So stale_entities isn't
        actually needed at all, and the calls to get entities aren't in
        most places either.

        ... also, this resolved the problem! It required a bit more
            intelligence around if/when a map is loaded, and I think
            that method is getting called every frame, so... I might
            want to find a way to lock that down. Later. :D