Currently we're using the internet archive.org servers during development of the cover art archive. After launch of coverartacrhive.org we cannot do that anymore, as we would be clobbering real data and introducing fake data onto archive.org servers.
As such, we need to emulate the internet archive part of things sufficiently to allow development separate from the production environment. I see three options here:
1. we ask archive.org to get a collection and a bucket naming scheme to be used for development and testing
2. we reintroduce support for amazon s3 to store cover art during testing
3. we implement a higher level abstraction which stores things locally
I'd prefer the third option, as it doesn't need a credit card and more importantly, doesn't need an internet connection.
Additionally, we need to:
1. add an admin/sql/CreateCAA.sql script of some sort, to set up the cover_art_archive schema for development.
2. add documentation on how to set up your development mb_server for cover art development (which probably would include CAA-indexer and coverart_redirect)