notes.txt

How document scope works..

While iterating over the document to populate event handlers, each level of the
document will create a new "scope".  The scope effectively mirrors the
document, and each level of "scope" is removed after parsing that section of
the document is complete.  The scope objects represent a hierarchy mirroring
the document structure.  Any level of scope can inherit keys from any of its
parent scopes.  Scopes are effectively singly linked, they know nothing of
their children, only of their parents.

Application of document scope..

The "new" action method is fairly special.  If an item is said to have a method
called "new" the actual method will never be invoked.  Instead, this is
indicative that the "new" keyword should instead be used to allocate a new
object of this name given the mapping from token name to constructor function.
This means that when an action which is "new"....


..THIS IS FUCKING COMPLICATED...

Population life cycle:

  populate() can be called before the configuration is available, but it will
  not do anything other than enqueue a population cycle.

  The populate() queue is emptied (by executing all pending population cycles)
  when the configuration file has been fetched.

  populate() will only populate the actual events for nodes who have actions
  that are already registered.  All actions that are not registered simply
  store nodes and a series of instructions for the given action.  At this
  point, registering an action will cause the node population to take place.

  registry-queue...
  action.name: [
    {
        scope: scope,
        type: action.type,
        node: node that matched,
        method: action,
    }
  ]

  while action.name is registered
    for each population in action.name:
        node.addEventListener(type, stub(scope.get(action.name), method))
        if (type == "load") /* Special case ? */


  events that have yet to be registered may want to enqueue items.  This might
  be overrated.

Overly Complicated....