tag:blogger.com,1999:blog-5514966116735614515.post6892143873366844271..comments2009-02-17T15:28:00.819+01:00Anders Nawrothhttp://www.blogger.com/profile/02538334376288340169noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-5514966116735614515.post-17234219364666247682009-02-17T15:28:00.000+01:002009-02-17T15:28:00.000+01:00@Tony: regarding: "No database or application can prevent you from recording untrue information, only logically inconsistent information."<BR/><BR/>They can be more or less helpful. Compare to pointers in C vs. Java. Java hides more of the implementation and thus prevents you from messing up things in some ways. I believe SQL forces you to think a lot about the implementation detail "key", where you rather should focus on your data. At the moment, SQL has defined our way to think about data models to such an extent that we actually have problems thinking about our data while not thinking in terms of tables and keys!Anders Nawrothhttp://www.blogger.com/profile/02538334376288340169noreply@blogger.comtag:blogger.com,1999:blog-5514966116735614515.post-45850835450842231082009-02-17T14:35:00.000+01:002009-02-17T14:35:00.000+01:00A common mistake: the word "relational" has nothing to do with "relationships". It comes from the mathematical term "relation" - see http://en.wikipedia.org/wiki/Relational_database#Relations_or_Tables<BR/><BR/>Also you say "But as long as a corresponding key entry exists there's no way for a RDBMS to tell if the application (SQL statement) got the relationship right!" Well of course there isn't! No database or application can prevent you from recording untrue information, only logically inconsistent information.Tony Andrewshttp://www.blogger.com/profile/16750945985361011515noreply@blogger.comtag:blogger.com,1999:blog-5514966116735614515.post-79323480104277478032009-02-16T21:07:00.000+01:002009-02-16T21:07:00.000+01:00My response: http://www.andrejkoelewijn.com/wp/2009/02/16/your-data-wants-to-be-free/Andrejhttp://www.andrejkoelewijn.com/noreply@blogger.com