Database Reference
In-Depth Information
3.7.4
managing Studio Processes
107
3.7.5
retrofitting Existing models
108
3.8
Sweeping Away the Dirt
110
3.1 introDuCtion
Every other Wednesday in my house is cleaning day. This would, on the surface, not
seem to be a significant occasion. however, at least it my home, it brings out an age-
old argument. you see, we prepare the house for the cleaning by cleaning it. The term,
I believe, is universally referred to as
precleaning
. he concept has oten escaped me.
Why do I need to clean for the cleaner? I pay her. She is excellent. I can guarantee
that she is more than capable of cutting through the clutter and focusing on the salient
details. nonetheless, I am compelled by unnatural force to acquiesce and “preclean.”
The unnatural force I am referring to is marital pressure. The entire discussion
around “precleaning” highlights a natural mental division between my wife and me.
We both have the same goal in mind—we want a clean house. however, to assume that
my wife and I approach the world of cleaning with a similar perspective is like stating
a duck and an otter have similar perspectives on the world because both get their butts
wet. Add to this two teenage/preteen boys and now the duck and the otter are joined by
a Beagle and a Labrador. All are interested in the water, but exist in completely different
universes.
Before I get in too much trouble, let me state (for the record) that I see the value of
precleaning. If you can cut through the clutter up front, you can more easily focus on the
real dirt. That it is not to say that my techniques do not work. If there is a stain on the
carpet, you can simply move the clutter and clean it. you can do that because you know
it is there. Conversely, if you do not know where the stains are, they never get cleaned
until the clutter is out of the way. Further, if you have ever dealt with boys, you know
that they are very good at hiding stains.
If you think about it, this concept is not dissimilar from creating an EPm (Enterprise
Performance management)/BI (Business Intelligence) system. you can approach the
solution from just the business requirements, focus on the reports, dig through the clut-
ter, and build the dimensions. The problem, of course, is that when someone else tries to
understand what you have done (no matter how good your documentation), they often
get lost in the clutter of data and metadata. Also, the clutter can often derail even the
most elegant solution.
The alternative to the focused approach is to remove the clutter and start with a clean
space. Both methods often yield the same final result. The important differences are
time, money, and the feasibility of actually cleaning things up. take my word for it, the
residue from sticky hands never comes of of the ceiling (at least not without paint).
The question we then must answer is how do you decide your approach? Do you
insist on the clean data and metadata before proceeding on a deployment or do you
start, knowing you have at least fixed it during the build in other projects and hope to
will meaning out of chaos on this one? It might seem like that statement lends itself to a
particular point on view, far from it.
In any implementation you have to balance available resources (time, money, people,
etc.) with an ideal situation. At the risk of using a generality, available resources and
ideal situations are perfectly asymptotic lines. They might get close, but they really do
not intersect.
