Ourmedia almost works

I have put the first link on the blog to a file I've uploaded to archive.org via ourmedia.org. Ourmedia.org offer free unlimited storage and high bandwidth for stuff you own the copyright to, or which is public domain. Ideal for podcasts. I'd like to do this for my personal stuff like Light Fission and also for the weekly Discovery podcasts. The only problem is that they use Drupal as a content management system, and its XML generation for podcasting is broken. So we've tried using dircaster.php instead. The problem is that ourmedia.org has a unique way of storing every single uploaded file. Instead of the traditional and functional system of a directory for the user, and then files under that, or under directories under that, ourmedia.org generates a NEW directory based on the Creator and the Title of the file. This means you can't use an automated system like dircaster.php to scan the directory for MP3 files and create the XML for us. There isn't a directory of just Discovery files to scan. There isn't a directory of just my files to scan. Instead there is a general downloads directory and every single individual file has its private directory under this. I'm stumped as to how to automate a remote generation of the podcast XML with such a lack of structure. The ourmedia tech guys assure me that whenever it is that they upgrade their hardware, the new server will have software that will generate XML for podcasting. Its a shame, because ourmedia.org offers everything else we need, including a way for any of the Discovery team to upload a show for podcast. We just don't have any way to automate a podcast XML file from these strange directory structures. Ultimately, if the ourmedia.org XML doesn't happen soon, I'm going to have to find some way of searching ourmedia.org by way of username and somehow extracting the unique URLs of all the associated MP3 and ASF files. Any advice or suggestions would be appreciated