Well.. I wasn't really asking whether or not you approve our testing methods. I was asking why it took more than an hour to retrieve the data. I found the reasons, and now know more about nhibernate. I mean no offense, but there really is context around this test that would take time to explain, but is not really relevant for my question.
But here goes anyway: In 99.99% of the cases, we(our customers) will load less than 50 rows. However, some(one) of our customers require that they are able to load all the data in a table, even if they'll never need or use it. They are paying our bills, and money talks. Therefore, we have two options: use NH normally 99.99% of the times, and create a stateless session with manual join of the objects for 0.01% of the cases. This does not make sense, as it'll be A LOT of extra work that'll hardly be used by anyone. Now, if NH could still load a lot of data within a few minutes, we could defend not having two seperate load strategies. This is why one of our many tests was to check if this was possible. It is, we're happy, and nHibernate replaces EF in our system ;-)
