On the front of your page you have indicated the following:

"After that there will likely be no weekly drops for a few weeks. Media Reader will begin using XBMC's database directly. Having looked at the database structure (besides the usage of the relatively inferior database SQLite), much of Media Reader's structure is almost identical. It seems only logical to no longer pursue the route of nfo files and rescans to have XBMC pick up Media Reader's information. This is inconvenient and a roundabout way for which I no longer see a necessity. Instead, Media Reader will write directly into XBMC's data repository."

While I can understand your desire to write directly to the XBMC database, I implore you to keep the option to create thumbnails and .nfo files for metadata when scanning media files with MediaReader.  While the XBMC database is a nice option, it limits itself to users who only use one XBMC-based device.  I for instance, use media centers in multiple rooms in my house, so having the metadata stored in .nfo files makes the scanning for new media impressively fast and allows all of my XBMC machines to update quickly and stay in sync.  When you do the reverse of this, which is having XBMC scrape all of the sites for metadata themselves, this slows down the database updating process drastically.  Plus if you have ever updated XMBC to an SVN version you will see that at times the devs make changes to the database which will require you to remove the DB and rescan your collection....without the .nfo files, this can take hours on a large collection.

Thanks,

TJ