![]() |
Welcome to Teacher's Pet, Inc.! You've been drafted to serve as the RDBMS design staff. Before I show you our data files, let me tell you about our company. Teacher's Pet, Inc. (TPI) offers highly specialized dog training. Have you ever seen the painting of dogs playing poker? Well, we trained those dogs! When owners of a TPI-trained dog say "Speak!" our dogs enunciate. When they say "Sit!" our dogs hire a baby-sitter. When they say "Beg!" our dogs hold up cardboard signs reading "Will work for food." TPI is a prestigious, exclusive, expensive, and (because we're so expensive) a dying company. Therefore, our marketing department —his name is Herb— has created a nifty ad campaign to boost business. In anticipation of the huge customer increase, the boss wants everything on computer. Here's how our data is currently being tracked on paper: |
|
|
|
These 3x5 index cards, by the way, represent the sum total of our current clientele. You can see why we need the new ad campaign. You can also see some of the problems we've had with data integrity. When our area code changed from 615 to 423, we missed Shadow's record during the update. You can also see that Nip and Tuck both know the same trick, but it's called "fetch" by one trainer, and "retrieve" by another. Also, the "Roll Over" trick was once mistakenly recorded as "Rollover." I'm also uncertain if the skill levels are recorded consistently. That is, does Nip play dead at 70% or is it 70 out a possible 300?
The salesman told the boss that a "Relational Database" would solve our problems. That's why we brought you in. Can you migrate this data to an RDMS? By the way, exactly what is a "relational" database?