DB migration

[hypnopotamus]

rather than seeding the database with data as the way to shape the graphdb soemthing more sophisticated is needed

something needs to be designed and implemented that can

• version the data model
• migrate existing data from one version to the next
• add / remove types from version to version
• still can seed data on a fresh database at the version matching the application

some of these things might be automatically take care of by the way a graphdb works, some might need explicit thought