I’ve never needed to use an OLAP cube before recently. I’ll admit that I didn’t see what all the fuss was about. But I came upon a problem for which an OLAP worked perfectly. Theoretically.
In practice (Excel 2003), I experienced error after error. I rebuilt the cube from scratch several times. And, in the end, was unable to get it to work on any machine except the one I created it on. And even then, it didn’t work the next day.
I’ll admit that I’m not an OLAP expert. I’ll even admit that I’m mediocre at best with regular pivot tables. So it’s possible that I just don’t know what I’m doing. But if I can’t figure it out, I weep for those with less Excel experience trying to make it work.
Despite my problems, I really, really wanted to make a blog post about it. I went through the tremendous trouble of creating sample data that was not only psuedo-random, but relational and normalized. No small feat, that. And just when I was that close…
I give up. I’ll try it on 2007 later, but for now I am 100% beat. You’ve squashed my soul, Excel. I hope you’re happy.