As a follow-up to last week's post, here is my weekly brain:

* There are some interesting LIFO/FIFO simulation tools out there, I wish I had some time to review a few of them. I may have to eventually, but for now it is a mere curiousity. Does anyone want to recommend one?

* The great space-time trade-off. Specifically I'm looking at the optimal balance between creating a transactional database with dimensional query speeds. What factors do I need to examine (obviously price points for physical storage, processing speed, etc.) My current assumptions lead me to believe it is cheaper to build a traditional mirrored data warehouse and ETL all data over. How would this be accomplished in a distributed system?

* Why does anyone ETL these days anyway? I think Bill Inmon is on to something with the whole DW 2.0 movement. Anyone care to explain how one would leverage metadata in a situation like this? It honestly is a new concept for me so I haven't given it much thought.