The nax_length gives counts in bytes. So if the text is utf-8, then we get incorrect result.
We have a code snippet where we check if the maximum length of the field and the max length of data stored becomes equal. This way we can check the potential truncation of data inserted