Setting the extent on a tile source now only means that tiles
won't be drawn outside that extent. Now the only way to specify
the origin of the tile grid is to provide a custom tileGrid. By
default, the grid origin is the top left corner of the
projection's extent.
The WMS used in the example does not send CORS headers (I think
not many WMS services do). It seems that the WebGL renderer
needs script access to the image data, so the example only works
with the DOM renderer in Firefox.
This ensures that maps with WMS layers with different extents
do not scale the tiles differently for each layer when the
default tileGrid is used. Note that it is always possible to
configure WMS sources with a custom tileGrid.
Telling the compiler that object keys are numeric causes more harm
than good (see === comparison that could never be met though the
compiler didn't catch it).
Never use goog functions without reading the source first. The
goog.array.sort method does nothing special sorting numeric strings
(so the zs array might be ['1', '10', '2'] here).
The render method always generates a map of all tiles that should be drawn
(called tilesToDrawByZ). This includes tiles at alternate resolutions, tiles
still loading, and tiles previously rendered. At the end of the render
sequence we can simply remove all previously rendered tiles that aren't in
tilesToDrawByZ. This provides an alternate solution to the problem described
in #53 and more.
The [URI standard](http://tools.ietf.org/html/rfc3986#section-4.2)
allows for URI without a scheme. Where http is used to load the app,
using a schemeless URL for the Bing metadata loads a script using http
- and the same goes for https.
This brings equivalent behavior when rotating in both renderers. Previously, offsets were maintained with style top/left. Now a single transform handles the translation and rotation. The scale3d function removes the separation between tiles. The next step for the DOM renderer is to scale in this same transform.
Now we use the browser's native dblclick event on pointer
devices, and we fire dragstart and dragend only if we're really
dragging, and not on mousedown/touchstart and mouseup/touchend.
We don't register two listeners for mouseup and touchend any
more. And instead of using the native click event on touch
devices, we react on touchend, to make clicking more responsive.
On pointer devices, the underlying native event of a dblclick
event is now a click event.
