Hi dphillips,

I appreciate your observation on the "nomenclature"! And as far as the article is concerned, thanks to Steve for his valuable comments.

Not sure why the approach is RBAR approach? In almost all cases, ETL stuff would involve transformation that work on row level data.

While I agree with you on having a normalized database structure, my requirement does not require much to do with the normalized data. If normalized, I would have to again de-normalize the data for generating a file as a feed to another application.