My investigation into the underlying data elements of the MARC21 record began in 2005 with an unpublished report that I produced for the Library of Congress that defined some possible steps that could be taken to prepare for “life after MARC.” At that time my assumption was that the next library format would be in XML. However, I stated that a necessary step in the process of transforming MARC21 would be to identify all of the data elements the format contains. In essence, a distillation of MARC21 into data elements would allow the same data to be carried in any future serialization. I wasn’t aware of how difficult a task this would be until I embarked on it later myself.