A database against the rules

In my previous blog I noted that I had restarted work on my database. In this post I'd like to talk a bit about database design and construction.

I'm guessing that most of my readers who are interested in data will fall into one of a very small number of categories. The first is a professional DBA who is either interested in seeing how other DBA's do things, or possibly here just for a laugh. The next might be someone tasked with throwing together an ad hoc database with a good amount of existing data, and presenting it in an easily digestible format without creating all sorts of dashboards,relationships, tables, etc. I'm certain there are many other scenarios which may pique one's curiosity, but my aim in documenting my journey is to stimulate an appreciation for and a desire for the application of data.

*For database neophytes: please tread carefully. A long time ago, I read something someone wrote that stuck with me: they said that a great database design starts on paper. So, unless you're like me and can "see" what your database will look like (and/or have built similar databases), design your database first, then start coding. 

Every database has a raison d'être. It's highly unlikely that anyone would wake up one day and say, "Hey, a database of all of the dry pasta in the house would be useful!" Obviously not... I think! (For the record, if you ABSOLUTELY need something to keep track of your pasta, a barcode scanner (or smartphone scanner app) and a spreadsheet would probably be a better solution.

When I was in high school, my appetite for data was whetted by my love of music. Personal computers were just starting to come on the scene, and I didn't even know what a spreadsheet was. Everything I did was done in ink on looseleaf paper. After a while, I graduated from just cataloging my music to staging popularity polls with friends and acquaintances. Quite surprisingly, most of them objected to my methodology, which was quite simple, and leveled the playing field. It worked like this:

A person could vote for up to ten songs and ten albums.I made sure to tell them to rank their choices by most favorite, then next most favorite, etc., so that the top song on a ballot would receive 10 points, the next 9, etc. My reasoning was this: just because 10 voters like Hey Jude, not all like it equally, and the vote should be weighted accordingly. 

For some reason, no one liked that idea.

But that, for me, was the start of databases. Stay tuned!

As always, I am hochspeyer, blogging data analysis and management so you don't have to.