- Data tables should be divided whenever more than one value is taking up a row or column
- Tables should be divided if there is a lack of data
- Different relationships and how to analyze data, such as many to many or one to many
Monday, October 28, 2013
Manga Guide
I continued reading The Manga Guide to SQL and after finishing Chapter 3 I learned that
Subscribe to:
Post Comments (Atom)
Good post! I would eventually like to replace the "Manga Guide to Databases" with "Getting Down with SQLite". Perhaps this is the time we can do it, but it requires:
ReplyDelete1. That we develop a shared understanding of the fundamentals of database design and normalization (Note: This is the easy part).
2. That we come up with some compelling high school student friendly example use cases that we can effectively use to tell the database story (Note: This is the hard part).
In an effort to move you along toward goal 1, I suggest that in addition to reading and rereading the summary sections from the Manga Guide, you look at the wikipedia page: http://en.wikipedia.org/wiki/Database_normalization
The important goal in database normalization is to avoid data redundancy and dependency. Regarding redundancy, imagine that I had a database of students with name and student number occurring in two different places. In one place I had you listed as "Aki" and in another I had you listed as "Sotirios". A mistake in entering your student number in one of the locations then makes it impossible to identify all records that pertain to you. This kind of situation is one of the things normalization is intended to avoid.