QUIZ and Suprtool use a fundamentally different approach to linking.

QUIZ relies on indexes to link data. For example, Access M-CUSTOMER link to D-SALES performs indexed reads of the sales dataset for each record in m-customer, and then it links each customer record to all the qualifying sales records. This linking strategy of one-to-many can result in numerous redundant disc I/Os, especially if the detail records are not stored contiguously and the file has a small blocking factor. QUIZ creates a subfile that contains all the fields selected from both datasets, including two copies of the link field.

Suprlink takes a completely different approach. First, the input and link files must be sorted on the same field. Then Suprlink reads the files concurrently and tries to link each record in the input file with the first qualifying record in the link file. As a result, Suprlink only reads each input file once using efficient NOBUF MR reads.

Suprtool can link on non-key fields, but it cannot perform one-to-many links. This prevents the output file from containing more records than the input or "driver" file. Unlike QUIZ, Suprlink does not repeat the link field in the output file.