# Thread: Basic ERD sematics question

1. Registered User
Join Date
Dec 2004
Posts
6

## Basic ERD sematics question

The "normal" appearance of an ERD, like I know it, can be seen on here:
http://www.utexas.edu/its/windows/da...m/erintro.html
on the bottom of the page, right above "Summary"

Now I stumbled several times over another notation, which I will try to illustrate:

country--(1,n)---<is in>---(1,1)--city

where <> is kind of a rhombus, representing a "relationship-type". The cardinalities in parenthesises representing the (min,max). So one could think that, one city is in "n", where n is at least 1, countries.
That, I think, is just plain wrong. It is of course just the other way round. "n", where n is at least 1, cities are in one country (asuming that a country has at least one city, which is quite a safe asumption).
What am I misinterpreting?
How would one design the city/country example in a "normal" ERD (with the first notation) ? Would it be something like this: country-1-----n-<city ?

2. Registered User
Join Date
Dec 2003
Location
Posts
710
There's nothing wrong with the example the left side shows the relationship of one country to many cities, the right side shows the relationship of one city to one country. In otherwords a country can have many cities, but a city of a certain name can exist in one and only one country.

3. SQL Consultant
Join Date
Apr 2002
Location
Posts
20,002
no, not "a city of a certain name"

rather, a city of a certain identity

what constitutes unique identity for a city is a different question, but i can tell you it is not the city name

toronto is in canada and it is also in the united states more than once

4. Registered User
Join Date
Dec 2004
Posts
6
Ok, thanks for the replies.

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•