I'm currently creating a digital library for 200000 scanned images of old books, in seven collections. Everything works fine on a small set, thanks to Omeka!
But Omeka saves each image in one folder, archives/files, and the server cannot support a large number of files in one folder. So we need to create subfolders. In fact, we would like that Omeka saves images with a path like that: archive / files / collection / dc:identifier / file_0001.jpg. Is it possible? When I hack the path of files in the test database ("archive_filename" in the omeka_files table) in order to add my_collection/my_identifier/ to the filename, Omeka continues to work fine. But it's only a hack... Any other possibility to do that?
In the same way, we prefer to keep the original filename of files and not a hashed one (this issue is similar to http://omeka.org/forums/topic/files-imported-are-renamed), in particular to be able to mass import them.
Last thing, we want the urls of documents have a permanent, readable and understandable name like: http://example.com/collection_name/dc:identifier (or dc:title). The same issue arises for url of collections. Currently, the "/show/" word in the url is useless for the end user and the id number has no meaning. Of course, we can use some purls service (http://omeka.org/forums/topic/using-unique-or-persistent-identifiers), but we prefer that Omeka manages them natively. Is there a plugin to do that?
I think that all these issues are close to each other.
Infodoc & Knowledge management