Most of our uses of source extent were cargo cult programming. The source extent was seldom and inconsistently used. Instead, layers can now be configured with an extent, and layer renderers limit rendering (and data requests) to the layer extent.
For vector sources, the `getExtent` method returns the extent of currently loaded features (this was the case before and after this change). For tile based sources, we will likely want to allow easy construction of tile grids based on an extent (this is not possible before or after this change, but could be added later).
Tile grids cannot currently be constructed with an extent (though we should perhaps provide a function that allows this - see 68815dca10 for an example).
This separates the action of requesting an extent to be loaded from the action of requesting cached features. The renderer (or any other consumer of a vector source) calls load to request a data extent. A `featureload` event fires when new features are loaded. The renderer (or any other consumer) separately asks for cached features given an extent. This vector source only loads features once, but this separation will also work with sources that make multiple requests for data in different extents.
This also removes the `data` option from the vector source in favor of a `features` option. Since we no longer have shared data structures for geometries, people can manually create features and pass them to a vector source. The `addFeatures` method is exported as well. This is used to add features to a source that don't have a representation on the "remote" (or server).
This moves the feature cache from ol.layer.Vector to ol.source.Vector. These are the minimum changes required to maintain the existing functionality and make tests pass. More refactoring to come.
