And, on memory (or address space) limited systems, another *right* way is to use the blob streaming APIs, starting with <u>[sqlite3_blob_open(...)](https://sqlite.org/c3ref/blob_open.html)</u> and proceeding with calls to APIs mentioned in the "See also" section there. SQLite can handle BLOBs that are much larger than many systems have enough memory with which to do the bind_blob() thing. The streaming APIs facilitate BLOB handling when large BLOBs are to be stuffed into or retrieved from a DB.