Currently, the same fact can get written to the dataset multiple times. To avoid this, we should use the epilog.js subroutine zniquify after adding facts to a dataset, and certainly before writing datasets to the database.
zniquify is called now before smth is sent to the server. This should be sufficient, i.e., it doesn't seem necessary to call this function in every subroutine that defines more facts.
Currently, the same fact can get written to the dataset multiple times. To avoid this, we should use the epilog.js subroutine zniquify after adding facts to a dataset, and certainly before writing datasets to the database.