[justinclift]

One of the senior devs (Martin) spent a loooot of time optimising the CSV import process. Many different approaches using sample data sets (from rough memory) of between 800MB and a few GB I think.