Text that is not valid UTF-8
What changes in Text that is not valid UTF-8 compared with MongoDB and what your application has to do.
This difference is of the Stronger guarantee kind.
What MongoDB does
It stores strings and field names that are not valid UTF-8.
What TeselaDB does
It refuses them with InvalidBSON (22) without closing the connection, like a subtype 2 binary whose lengths do not match.
What your application has to do
Validate text in your application before writing it.