This will likely be a new (3rd) index set including these index types.
We also need new pipelines with collation.