Greg writes:

> In case this touched on my comments on TPC-C, I wanted to clarify that I did
>  in fact mean using Image, not Oracle.

One of the things about the TPC standards is that they basically want to use
the same software on all of the machines so that they have a consistent test
by which to compare the various platforms. The only exception to that
statement is that the various RDBMS manufacturers use these same tests to
compare their various products on a a single test machine.

However, as I've said here many times before, on a reasonably real-world
business test, such as TCP-D, I am absolutely sure that an IMAGE database,
populated with the standard test data, and reported on using a product like
QueryCalc (which would be ideal for the test data), would "blow the doors
off" (a drag racing term) any equivalently loaded RDBMS accessed through SQL
calls.

I would be pleased to perform such tests if it weren't for the futility of
the exercise. If I understand the rules correctly, the results wouldn't be
accepted, no matter how honestly and fairly they were conducted.

Wirt Atmar