I have an application that has to insert about 13 million rows of about 10 average length strings into an embedded HSQLDB. I\'ve been tweaking things (batch size, single threade
