Explain how the DB is locked

[szarnyasg]

There is a short FAQ entry about this but it is rather vague https://duckdb.org/faq#how-can-multiple-processes-write-to-duckdb

1. Explain why does this limitation exist? Compared to SQLite (which does not have a cache) or Postgres (where you cannot launch multiple Postgres instances on the same database file).

2. How do your design patterns need to change --> that's a guide e.g., "double-buffering style" for locking. Here's how you do this in Python.

[szarnyasg]

This should go under the new Concurrency page.