
NEWS.md
A bug in separate_() is corrected: when into= contains NA and the data frame contains other columns, insertion of new columns failed.
On the contrary to tidyr::separate(), separate_() does not allow any more all NA values for the into= argument (meaning we want to keep none of the separated columns). It raises an error with a tip to use select_() instead to eliminate a column from the data frame.