While this might be an older thread, I would like to add my agreement to it.
This blog has done some interesting work:
There needs to be a consistant, supported, way of databinding, not only the properties on the map, but the data on the map too.
I would love to be able to bind a collection of View Model objects to the map as a new layer. And then have the collection's "visuals" and its location, determined through templates. I really dont understand why the VE Map
Control didnt have this from the get go. There should be a Layer template and a "Point of Interest" template.
Then when I want to add a layer that represents earthquakes, I hit the USGS rss feed, get my view model of earthquakes, have a control template that represents an earthquake, create a layer, add the collection to the layer, add the layer to the map, and
its done. The control template would use the GeoRSS point lat and long from the rss feed to communicate the position of each quake to the map. The magnitude would be something that was shown by the control template converting between a magnitude
and a width. I could let a designer have all kinds of fun with how they show the quake.
I can do this in code, but it would be real nice to be able to use MVVM with an established, supported pattern.
Thanks again for all of the great work that you guys have done on this site.