These innovations provide powerful facilities for dealing with a very difficult problem - missing information. Using FirstSQL's capabilities, database users can confidently utilize the power of the relational model in this area.
With two types of nulls, the representation of missing information can more accurately reflect the real world. Additional manipulation tools allow the users to more effectively utilize their databases and probe more deeply into the meaning of the data. FirstSQL's robust implementation insures the integrity of the data and the results.
The question is - does your database handle missing information properly? As C. J. Date points out, avoiding the use of nulls in database columns will not suffice, since nulls are routinely produced by SQL operations such as aggregate functions and outer joins. Besides, the issue of missing information can be rarely be avoided, and ad-hoc solutions are even more prone to produce erroneous results than current RDBMSs.
The relational model has good solutions for missing information, but does your DBMS implement them properly? Test your DBMS with the queries (Examples 3, 4, 5) in FirstSQL Test Drive against the correct answers produced by FirstSQL.
Return to Contents Page: Handle Nulls Properly? Return to Can your Database do this?