Cover the main differences between the KSPFetchedResultsController and NSFetchedResultsController

[konstantinpavlikhin]

• There is an important difference in a way KSPFetchedResultsController merges the changes from the NSManagedObjectContextObjectsDidChangeNotification. The chosen way is better for NSTableView pairing, but may be a little confusing for the NSFetchedResultsController users.
• KSPFetchedResultsControllerDelegate is notified not only about changes that already took place, but also about changes that are going to happen.
• KSPSectionedFetchedResultsController uses NSObjects as 'names' for its sections as opposed for NSStrings used by NSFetchedResultsController.
• KSPFetchedResultsController uses plain indices instead if index paths used by NSFetchedResultsController. Again, this is better for NSTableView pairing. Because of this the KSPSectionedFetchedResultsControllerDelegate method that is called on object change has completely different signature.
• When reporting sections changes KSPSectionedFetchedResultsController has three possible events: section insertion, section movement and section deletion. Again, delegate method signature is a little different, and a separate enum (KSPSectionedFetchedResultsChangeType) is used to define possible sections changes.