Claudio wrote:

> You may argue that's too bad that he dropped IB in his
> list. Imagine how
> worse could be if he does a review if IB given his current writings! Also,
> keep in mind that the MGA design of IB doesn't make it appear as fantastic
> in single-user, raw benchmarks. You need a real case => utopia?.

Actually, I gather Postgresql has it's own version of MGA transactioning
called MVCC. I'd be interested to hear a _technical_ comparison of the two
designs and/or implementations if anybody knows.

TTFN,
Paul.