Namespaces in models

Many ERP systems allow its user to write applications on top of them. Applications may store their stuff in the system's database, which means that each application has its own set of entities. If we're to write such a system using Persistent, how can we handle a case when two unrelated applications create entities with same names?

In other words, if we have two Haskell modules M1 and M2, both of which containing something like

share [mkPersist sqlSettings, mkMigrate "migrateAll"] [persistLowerCase|
User
   name Text
|]

then running

import qualified M1
import qualified M2

insert $ M1.User "User1"
insert $ M2.User "User2"

will place the data into the same table.

Would it make sense to implement some sort of namespacing in Persistent? If yes, what approach should be taken?
I'm thinking about taking a fully qualified module name (like, packagename-1.2.3.Database.Persist.Foo), computing a hash based on it and appending it to the table name.

Start with the Persistent schema declarations using share, mkPersist, and mkMigrate, and compare the generated entities from the M1 and M2 examples. Trace how entity names become table names and how migrations represent them. Done would require an agreed namespacing design that keeps unrelated same-named entities separate and works with migrations.