For a long time I’ve never thought of Data Modeling as a full time job. For me it was more a step along the way to getting a database built. A tool to fiddle with data concepts until the data fell into place so that the database could be built and the applications built on top of it.
I’ve always known that E.J. Date was onto something. That the Normal Forms were a great discipline for managing data structures and it’s relationships. I think the Entity Relationship Diagram (ERD) is what got me started in relational databases. The way you can take business information and practices and break them down into components is, for me, fascinating. Of all the programming and flow charting and other diagraming technics the ERD is most appealing to me.
In highschool I took technical courses (shops) like woodworking and automechanics … Something that is sadly lacking in my own children’s education (but I digress)… In grade 11 and 12 I took architectural drafting. Drawing with pencils and straight edges, learning perspectives and imagining and depicting three dimensional objects displayed in two dimensions on paper.
I could have been a draftsmen. For me there’s something very aesthetically pleasing in drawing the lines on a page. Creating a plan for something that can be built from drawings. Calligraphy without words.
Over the years, computers have made paper drawings and draftsmen obsolete. Computer graphics have come a long way since the late 1970s. But I still like the idea of creating plans that can be built. Whether they’re for a toolshed, a desk, a stage set, or an enterprise data warehouse. The idea of drawing a picture and then making that picture come to life. That is cool! And has huge Karma, for me.
Well I woke up this morning thinking about a data model I’ve been working on and how much I enjoy drawing the lines on the screen. It’s not the pencil on the paper scribbing the line so much as the line itself and the meaning it takes on when it connects to other lines and shapes that convey more meaning.
Kind of like cave drawings. Telling a story with pictures that speak their own language. Communicating ideas – how mundane, how human.
Blog posted here.
Your article is really nice and I truly enjoyed reading it.Waiting for some more wonderful articles like this from you.
I always remember Harvey saying “I am simple man, you have to make it simple for me.” Really quite a challenge for people who have the “curse of knowledge.” Sometimes having to explain things to a complete outsider helps even though it takes patience. The overall concept must be simple if you can explain it simply. The trouble with make things simple is often very complex and harder than making things complicated.
Easy to say but really an important principle. If it cannot be explained to Harvey, I expect it is too complex and not good design.
Now that is putting the cat among the pigeons.
In the last 72 hours I’ve spent about 14 moving enities, connecting them with relationship and simplifying two divergent models down to one. In order to draw the boxes and lines and have them make sense there is no question that I have to have prior knowledge of the business (either personal or gleened from subject matter experts) and a modicum of modeling skills . There is no question that trying to make the model simple to understand is a BIG challenge. It’s just so easy to keep cramming more boxes onto the diagram!
I think data modelling goes way beyond pictures. One think that is very important is the relationship between the little little boxes. They provide a visual image of complex relationships.
The models make visual some very complex ideas and makes it possible to communicate and discuss.
However the genius of data modelling, I think, is the ability to make them simple. Making the model simple is much more difficult than just modelling relationships slavishly. Any machine can model relationships like reverse engineering. Good models are much simpler and easier to understand.
Now what do you think of that, Graham?