We manage ERP systems with larg schemas (150.000 - 200.000 objects) where only a small proportion - the custom objects - should be under source control / schema compared.

To enable a decent use of Schema Compare for these systems, we would need to be able to compare schemas partially based on naming convention. Multiple assignments per schema should be possible, e.g. we want to match all objects for schema ABC matching the naming convention EE% and FF%

It seems materialized views are not included in the compare. Or am I missing something?

If a column is renamed, Schema Compare for Oracle sees this as a drop and create, so during the table rebuild data is preserved asides from the “new” column.

Could you make this work similarly to SQL Compare and execute a rename instead?

It would be useful to support files with different extensions, either by allowing to configure ignored extensions (all other will be treated as sql) or by allowing to configure extensions for specific types of objects.

I have a schema where synonyms are not being compared correctly. In particular this is a synonym that refers to a remote object. Is this a known limitation of Schema Compare?

I must prevent to drop any tables, because the database may contains other tables for another application.
And yes , they are in same schema.

I need to deploy database changes on multiple databases, at client site, from a single snapshot.

When I do a schema compare that has added a virtual column the alter table command that it generates has a bug in it.
The statement it generates is:
ALTER TABLE vendor_portal_user.vendor_contacts ADD (dbupper_username VARCHAR2(30 BYTE) GENERATED ALWAYS AS (UPPER("DBUSER_NAME")));

Which results in error:
ORA-54015: Duplicate column expression was specified

The correct syntax is:
alter table VENDOR_PORTAL_USER.VENDOR_CONTACTS add dbupper_username as (UPPER(DBUSER_NAME));

when migrating indexes in schema compare,bitmap indexes on a partitioned table fail - local clause missing