> Also - plan to drop the TIBOTables when you clean up your migrated
> application and learn the mysteries of SQL. TIBOTable is transitional -
> it exists only for compatibility with the Paradox model you are finally
> abandoning, to ease the migration from the BDE.

You don't need to be in too big of a hurry to get rid of your TIBOTable
components because TIBOTable automatically makes use of horizontal dataset
refinement if you have supporting indexes in the database.

All the other components you have to manually configure in order to make use
of this capability, which I'm sure would be a tremendous amount of work that
may not be necessary.

What horizontal dataset refinement does is have your dataset act more or
less like a scrollable cursor for the entire table that efficiently uses
indexes to jump from place to place without having to fetch all the records
of the table into the buffer. When you perform a Locate() or jump to a
bookmark, this is performed on the server and only the matching records are
brought back.

TIBOTable encapsulates a significant amount of logic so that your migrated
application will continue to deliver paradox-like performance even though
you are dealing with large tables and potentially many more concurrent
users.

You will perhaps want to get some help tuning it properly so let us know how
it goes. You need to make sure you have indexes created for both ascending
and descending access to the table for whatever you are ordering it by.
IndexFieldNames is likely the property you will want to make use of.

Kind regards,
Jason LeRoy Wharton