Total FAIL on the nHibernate code. Classic mistake of someone who doesn't know anything about nHibernate. NHib has 1st level caching and so it's not appropriate to put it into a tight loop without either first disabling the cache or clearing the cache every few loop iterations.

Like I said that wasn't the point of the exercise to make it quick.

But with your help we can make it quicker.

I remember a similair post few weeks ago. It also made the same mistake as mentioned by chadmeyers. I do wonder what people are trying to prove with these tests... When comparing sprocs,sql with or mappers they always tend to forget that NHibernate does much more then just performing selects and crud actions.

Well JjVv, I think you should read the conclussion first then, I don't care about the numbers. But people keep reading the numbers and then draw there own conclusions. I wonder why?

they read the numbers and draw their own conclusions because you put the numbers out there. (which we know need serious tuning) .....

you tested a non realistic use case.... you blogged it as a performance test....you did it without expert knowledge of each tool (or of VB.Net it appears) and you then conclude that the results don't matter? That doesn't fly... the point of such a test is that for whatever reason you care about the results. Otherwise the test and the process are pointless.

you didn't care about the results... so you should never have executed the experiment... but thanks for putting more fuel out there for the anti ORM folks. just because you post some data doesn't mean it won't be used out of context.

Yep, I can prove anything with numbers it's called statistics. I can prove the exact oposite if I need too. But do I really have to? If you can't draw your own conclusions then you shouldn't be reading the post.

And what people think about my coding capabilities doesn't bother me one bit.

Oh well, I knew this was going to happen. But we live in a free world.