Now the application no longer needs to care about the tile origin, because the CacheWrite control modifies the url if the CORS image loading is disabled and it is from a different origin. This only requires OpenLayers.ProxyHost to be properly configured. Also local storage keys use the original url instead of the proxied url, to make the CacheRead control work without proxy settings.
No deferred exceptions are thrown any more. Instead, OpenLayers.Console is used to show an error message for security exceptions.
We now check for OpenLayers.Tile.Image, because other tile types (e.g. UTFGrid) are not supported (yet).
To make the same origin handling in the CacheWrite control easier, OpenLayers.Request now exposes the same origin logic from request.issue as a separate function, so it can also be used by other components.
The zoom control allows for zoom in/out links that can be styled with CSS.
Note: This change was originally captured in #269. It involves nice additions thanks @ahocevar. The changes were unintentionally merged and then reverted with fb3caf1561, so the history of commits is not immediately apparent (though still likely there somewhere due to the magic of git).
Previously, a map could not be destroyed if it included a permalink control without an "element". In addition, a permalink control could not be destroyed if it didn't have a reference to a map.
Having to call two methods to get complete feature information (id and data) is cumbersome. The `getFeatureInfo` method returns an object with both feature id and data.
The tile now has responsibility for resolving feature ids and fetching feature data given x, y pixel offsets with getFeatureId and getFeatureData methods. The layer has corresponding getFeatureId and getFeatureData methods that take a map location, lookup the appropriate tile, and delegate to the tile for the rest of the work.
I think it should be the job of the layer to retrieve data for a given location (instead of the control). The first part of this change creates a `getData` method on the layer and updates the control to use this method.
The second part of this change removes the assumption that the data returned will be an simple object representing feature attributes. The UTFGrid specification doesn't say anything about the structure of property values in the optional data member. The examples given in the spec use string values. The default callback previously assumed that the data could be rendered in a two column table. I think it would make more sense not to make this assumption. With this change, the user must always provide a callback to do anything with returned data.
When the list of event types became unconstrained in 501b42228a, we lost the documentation for events that are triggered. This change adds the list of events triggered to the API docs for events properties.
