• The only bad part about Charles doing the testing is having the proverbial mouse guard the cheese.

    Actually, Jeff, in this case, I look at it as "put up or shut up" ;-).

    Regardless of the author's original computer specs (although that made me curious too, thanks for asking hiim), I'd like to see the vendor run the code in this article, on any machine or set of machines, plus his own version, and provide all stats on a level playing field.

    Here's a case where we *know* the test was not vendor-designed. If Charles runs exactly that code, plus his, in like conditions (and I invite you to add conditions: start sql server up cold? anything?), we should really learn something.

    >L<