that's a rather provocative question you're asking there.

What I find interesting in this analysis is the difference in the amount of inserts between the different implementations. No, I did not download the code and test it myself, but I'm a bit worried this test really is identical for all implementations. The amount of queries is definitely a measure that is important and varies between implementations. But the amount of inserts.... I can't think of why one implementation would perform significantly more inserts than another. Either there was also a few thousand deletes too, or the implementations wrote completely different data into the database. And in that case I'd be more worried about correctness than performance.

Dirk