You can set the delete command to 'NONE' on the snapshot commands tab. The
downside of this is that all the deletes will be prevented, not just the
archiving ones. If you are only logically deleting records in normal working
then this will not be an issue though.
HTH,

HI paul,

Thnaks to all you reply..
Let me explain you a bit the environement we are working on , then you might
have a great idea to solve my issue...

We have an industrial application(developed with VB) which is in charge to
collect measure data from different running machines. Those data gets
collected and archives in a local database. Then if our customer would like
to centralize those data, we install him an export solution into an SQL
server data base £(develop with VB). Today as a local database we are using
Microsoft Access ( which is not suitable anymore), and periodically or based
on user action, data gets sent to SQL server through ODBC ( really bad).
After a while, exported data are them deleted from lcoal database.

Today we are changing all of this, the local database will be replace by SQL
server 2005 satandard edition, and in case our customer woul like to export
those data, we can ask him to provide an SQL server 2005 server on which we
configures our tables. the advantgae here will be that we can manage the
export of data without having to develop anything as SQL server can take care
of that. Thats why I was thinking of replication.

Then I make some testing using the transctional, and pull method.. But as
you emntioned to me if data gets deleted from the source there will be
deleted to the destination. And this absolutly what I would like to avoid. I
simply just want to archive them automatically.

BAsed on that requirements, shoudl I better set the replication as snapshot
instead ?
Does deleted dat will be also deleted ?

regards
serge



[quoted text, click to view]