Enhancement Request for "Compare User Objects" tool

[Patrick Holmes]

I use the "Compare User Objects" tool all the time. However, we also use Oracle's Data Masking (DM). DM has a really annoying bug that changes the precision of NUMBER type columns. For example, we have a NUMBER(9) for SSN in a table and a NUMBER(38) for PHONE_NO in the same table, in Production. We often create a Test database from a copy of Production and then mask it. But the DM changes the type for each masked numeric column so that it is changed to just "NUMBER". Then when I compare TEST to Production, there are differences that should not be there. Oracle won't even acknowledge the bug per se, simply saying that it is "expected behavior".

If I try to change Test to match Production, i.e. change the type from NUMBER back to NUMBER(38), I get a an "ORA-01440: column to be modified must be empty to decrease precision or scale" error, of course.

If I were to change Production to match test, our DBA believes that this could easily affect performance due to chaining and increased storage.

What would be great is if you could provide a comparison option that was a checkbox for "Ignore number precision", so then your comparison would treat NUMBER(38), NUMBER(9) and NUMBER as all the same type. The option would default to unchecked, but I would use it when comparing with a Data-Mask'ed database.

I know this is a very odd request, and would not normally be a good idea. But if you had this feature, it would make our lives a lot easier.

thanks,
pat

 

[Marco Kalter]

I have added this to the list of enhancement requests.

 

[Patrick Holmes]

Have you given any more thought about this request? The issue is enough of an annoyance and time-waster for me that I will pay you US$250 (in addition to our annual license fee) if you make this change.

I expect that you don't normally assign priority to enhancement requests by having people pay for changes, but I thought to myself--why not? Maybe instead of having people vote for enhancement requests, you could have people *bid* on changes? It's just a crazy idea, and maybe users would resent being asked for money for enhancement requests they feel should be free, so it's probably a bad idea in general. However, in this situation, I would be happy to pay for the change because it would actually save me time at least once a month, for the rest of my programming life.

Again, I realize that this is an unusual request. It is only necessary because Oracle's DataMasking has this unusual design flaw. (They won't acknowledge that it is a bug.)

Regardless of what you decide to do, I'm still a very happy and enthusiastic fan of your product! Keep up the good work!

thanks,
pat holmes