several db ?

olivier vidal_olivier at yahoo.fr
Wed Sep 29 09:46:42 CDT 2004


Valentina 1.90 RB 5.5.3 Mac os X 3.3

Hi Ruslan and list,

I makes a program which manages several address books (name, first 
name, address, cp, city, telephone, mail....).
The customer should be able to make comparisons between address books, 
in a simple and fast way.
Address books can have many addresses (they are intended to make 
mailings). (500 to 30 000 adresses per adress book)
It is important in this application of power to compare (to compare, to 
deduct, to make requests on the same field), for example, the persons 
in the adress book "customers" and the persons of the adress book 
"prospects".
( For example a customer would like to know all the persons of the file 
customer AND of the file prospects who have the same zip)

I hesitate in the choice of the structure:

- A adress book = a db?
Advantages: very fast treatment inside the adress book. Number of 
infinite possible adress books.
Inconveniences: treatments between various adress books impractical and 
not expresses.

- A adress book = a table?
About the same advantages and the inconveniences as the previous 
method. Limited number of adress books.

- All the adress books in the same table?
Adress book would be identified in the table by a field byte.
Advantages: treatments between various adress books fast and flexible.
Inconveniences: additions and updates slower. Treatments in a single 
adress book averagely fast.
But I have the impression (thanks to the speed of Valentina) that this 
solution is ideal if all the adress books does not exceed 100 000 
addresses.

According to your experience, that think of it you?
The last solution seems risky but if the customer manages on average 
1000 at 50 000 addresses, it is optimal?

thank you very much

olivier



More information about the Valentina mailing list