A feature request: A differential version

[dustyghost]

Check the DB for migrations already run and compare the physical migration files with the db. If the object is in the DB, but not in an already run migration file, then add it to the next migration. Then generate new migrations for only new columns, keys\indexs ect....

Object drops could be done the same way. If they are in a migration file that is already run, but not in the DB any-more, they need to be added to the next migration as a drop.

This is a massive task I know, but it would save a ton of time if we had that! :)