Hi everyone,


Staff
StaffId *
Name

Department
DepartmentId *
Name

StaffDepartment
StaffId
DepartmentId

However he would want to do this:

Staff
StaffId *
Name
DepartmentId (Nullable)

Department
DepartmentId *
Name

.

I dont think explaining will help. You are a smart person, and i may suggest a different solution (it does help, if used the right way):

1. Go to your boss' office and say hello to him.
2. Pick a heavy book from his desk.
3. Hit him on the head with it.

In other words, dont waste your time explaining rather simple things.

[My Conclusion:
Database design always follows business rules.
1) If one staff member can be assigned only one department then your boss is right coz ur design allows multiple departments to one staff member (violation of business rule)

Hi

First off good luck convincing him that you are correct. (You are correct)


Relational databases should contain minimum amout of repeated data for following reasons.

a) Your bosses solution will need more disk space which meabs records take longer to load etc. Try calculating extra disk space required. You know how many records and field length allocated for department.

b) Multiple department entries (your bosses method) will lead to spelling mistakes will lead to bad reporting, etc.

Your method will allow for the department to be entered once. One possible spelling mistake, one place to update, etc.

Your method leads to easier reporting, less errors, less disk space used, quicker running.

Let us know how you get on please.

Denis

One poster had mentioned that doing it your way means no code change when/if 1:M is later needed, but that may be untrue if the business object is designed to expose a single entity for department as opposed to a collection (which it should if the biz rules call for that). Not to mention any UI differences such a business rule change would cause.