primary keys and replication

Hello All,

I am attempting to write a point of sale program
bassed on mysql, and ms access for the front end of the back office. since I will have two stores each with there own server with a copy of the database and adding records to the (say transaction table).

Is there some way to get a unique key so the transaction table can be replicated. or is it just better to duplicate the tables.

I read something about creating a two column primary key, although I am not sure if microsoft access can handle it.

Any thoughts would be appreciated.
Peter
Many thanks

You can define composite primary keys in MS Access. Which database will be the `main` database? What sort of `issue(s)` are your confronted with?

Ok,
Then if I understand correctly all I have to is for each table that could be modified in a different location, is to compose a primary key, which is a composite of locationid and an autoincreament. Then I don't have to worry about replication creating duplicate primary keys. I had planned on using ysql with innodb tables as the main database, and using access to enter the supplier purchase orders, and querys for daily sales etc. For the clients (the pos terminals in the shop) I was going to use qt and c++.

But I am very open to suggestions and advice

Many thanks
Peter

What issues exist to using MySQL for both the front end and query tool? This would eliminate you having to integrate to distinct backends.

I actually thought that using the ms access as the backend gui would be quicker. But to be honest the more normal the database is the harder it seems to get ms access to do the right thing. So perhaps I am with you and I will drop ms access.

Thanks again
Peter