This talk traces the history of data streaming algorithms for the cardinality estimation problem from the 1980s to 2010s through algorithms designed or studied by Philippe Flajolet, and the origins of the world-famous HyperLogLog algorithm.