>Seriously though I think that you and Jim really are a lot closer than most might take. In Jim's case I know he has the real life experience which backs up his 'take' on normalization though he can certainly speak for himself. I've also heard him say that if you need to denormalize to make something work just do it (Am I remembering correctly Jim?) but always plan for normalized approaches.

Doug,

You are right , I have said that and do say that. But I am very very careful about when and to whom I say that. It is much too easy for new folks to interpret that as saying that normalization is of not value. When I do say that it is in the context of having a fully normalized design in place already and wondering about to solve some specific problem that has arisen. The key point being that"a specific problem HAS arisen (not is anticipated or seems like ti will show up)" and "a fully normalized design has already been accomplished".