URGENT: Primary key vs Unique Index

I am setting up a symmetric replication environment. Most (but not all) of the tables in the schema have unique indexes instead of primary keys.  For replication manager to work properly, there needs to be a primary key on the table. I am tempted to just write SQL statements to convert the unique indexes to primary keys. What problems would I have by doing this? What is the difference to Oracle?

I need to complete this by Monday, so quick responses are greatly appreciated. Also, please reply to my email address rather than the newsgroup/list, as I'll be busy with Oracle and can't wade through the messages :).

TIA, jjr_at_healthfare.com Received on Fri Feb 14 1997 - 00:00:00 CST