Because the results are not (scientifically) reproducible. (Well, they are for that one implementation...)
Because data storage is part of the Pick data model, we can be confident that benchmark results are valid across all (similar?) implementations. Because relational forbids knowing anything about the implementation, any benchmark is valid for that one configuration only. Cheers, Wol -----Original Message----- From: Geoffrey Mitchell [mailto:[EMAIL PROTECTED] Sent: 16 July 2007 19:47 To: u2-users@listserver.u2ug.org Subject: Re: [U2] UniData 7.1 vs. MS SQL 2005 performance > Read Codd and & Date's rules. Can't remember which, but one of them > says "the database user is not permitted to know how the database > stores the data". In other words, empirical testing is FORBIDDEN. > Seeing as empirical testing is *the* "sine qua non" of science, > relational databases are, BY DEFINITION, totally unscientific. > > So, if you want a benchmark, relational database theory explicitly > says "No way, Jose!" !!! > How does not being able to know how the data is stored prohibit empirical testing? As long as you can provide inputs to a system and retrieve outputs you can perform empirical tests. -- Geoffrey Mitchell Programmer/Analyst Home Decorator's Collection 314-684-1062 ------- u2-users mailing list u2-users@listserver.u2ug.org To unsubscribe please visit http://listserver.u2ug.org/ ------- u2-users mailing list u2-users@listserver.u2ug.org To unsubscribe please visit http://listserver.u2ug.org/
