I use schemas to make separate logical divisions within a database to partition off common functionality. I’ll also create a file group to hold data for some schemas. One I have is “ref” for common lookup values that are very stable (man of schema = ref file group = Reference). Another schema I have is “io” – guess what I do with that one?

I also use schemas for security purposes (create role and assign roll access to objects based on schema). Schemas make it simple to have DRI between common objects (list of states, departments, employees, etc.) and application specific objects (accounts, loans, sales orders, etc.).

I can easily see how to apply this in a datawarehouse (fact, dim, stage, load, rpt, etc.)

When it comes to SSMS, I like the idea of having a flat view of all objects by type but having a schema organized listing would also be nice. How about a user option to permit a filter at the database level that cascades down thru all object types?

@jeff Moden – I don’t want to shill for a product but HyperBac (recently acquired by Red Gate) allows you to connect to backups via an ODBC driver so you can recover specific items from a backup. We use their product for backup encryption/compression and have used this feature several times. One of nice features is that the backup is smaller so you can avoid the T-Byte connection. I guess you can mark me down as a fan on this one.