PivotViewer is the latest tool to come out of the Microsoft Live Labs Pivot project. Pivot is based on collections, which are sets of data where each item has an associated image. A pivot item has attributes, similar to properties, called facets; and facets have facet categories. Facet categories are used to filter and sort the data.
More complex Pivot data sets have several linked collections, or dynamic collections which are generated at runtime as a query result. This is necessary if the size of the data set is very large or even unbounded. You could create a web search, for example, that returned a pivot collection.
Once you have created and hosted your Pivot collections, most of the work of displaying them is done by the Pivot client. There is a desktop Pivot client, which is Windows-only; but the Silverlight PivotViewer is more useful since it allows a Pivot collection to be viewed in a web page. The client (or control) does most of the work of displaying, filtering and sorting your data, including a user-friendly filter panel.
PivotViewer also makes use of Deep Zoom, also known as Seadragon, which lets you view vast images over the internet while downloading only what is needed for the small section or thumbnail preview you are viewing.
The result is that a developer like Azure Technical Strategist Steve Marx was able to create a PivotViewer for Netflix with only about around 500 lines of code. This kind of product selection is a natural fit for Pivot.
I was quickly able to find the highest-rated music movies in the Netflix Instant Watch collection.
Starting with the full set, I checked Music and Musicals and then set Rating to 4 or over.
It seems to me that the strength of Pivot is not so much that it offers previously unavailable ways to visualise data, but more that it transforms a complex programming task into something that any developer can accomplish. Microsoft at its best; though of course it will only work on platforms where Silverlight runs.