On Mon, 2006-08-28 at 16:02, Joshua D. Drake wrote: > Tony Caduto wrote: > > http://newsvac.newsforge.com/newsvac/06/08/28/1738259.shtml > > > > Don't know the validity of this dvd order test they did, but the article > > claims Postgresql only did 120 OPM. > > Seems a little fishy to me. > > > > This has got to be a complete joke. I'm wondering if the source code is available. My guess is it was full of MySQLisms and the postgresql "port" was written without indexes, no transactions, and relied on running dozens of queries that postgresql could have more efficiently done in one query or in one transaction at least. I wonder if one couldn't get ahold of the code and "fix it" to see what postgresql could actually do. We have the same problem at work. RT is a great little program that supports several database backends. But it's dumbed down to make MySQL 3.23 happy and fast, so it's kinda a dog on PostgreSQL, even though it's ultra stable. Seems the newest release is finally catching up on PostgreSQL.
