Picturepark 8.6 introduces a new SQL/NoSQL search hybrid that we believe will offer customers the best of both worlds. There's no question that NoSQL search indexing offers impressive performance and options with regard to search results ranking. But few enterprises are willing to do without a proven SQL database when it comes to storing their data.
Picturepark was already based on SQL Server, but our previous search technology was showing its age. The solution we came up with was to graft a NoSQL search layer atop our SQL Server database. The result is extremely fast search performance, especially for instances with millions of assets or thousands of tags. Even better, because we're now using our NoSQL layer for all of our UI caching operations, users see a general performance improvement across the board.
Virtually no UI changes were needed to accommodate the new search technology, so no user retraining is required. (A few search options were removed that are no longer required.)
The new technology also enables Picturepark administrators to rank each metadata field so that when matches are made, Picturepark can determine which assets are most relevant to the search query.
Thanks to Picturepark's SOA design, this technology swap was relatively painless for our engineers. This underscores the value of having such a flexible (and forgiving) software architecture under the hood.
We've also swapped out our video playback technology in order to provide a better user experience. And we've added the option of creating video cue points (bookmarks) that users can use to tag and jump to video locations.
