ED HAYES

I communicate with Oracle support regularly. Here are some of the recent response I have gotten from them.

Example 1

This is an example from a severity 1 SR.  We are unable to drop a picked LPN/pallet in the warehouse.  The SR was originally severity 2, but because we had received no constructive responses for a shipment that wouldn’t ship for 3 days, we escalated it to severity 1.

From Development
———————————–
The LPN #####  is showing in context 5 so looks like the cartonization_id
stamping of this using custom program might have caused this. The LPN that
got used for transfer is in status defined but not used. At this point the
best step right now would be to unload and remove the allocations from the
Transact Move Orders form and backorder the line. A datafix would essentially
do the same and then require the customer to re-release/pick load and drop
the lines.
.
Pl. try the action plan to backout and cancel the allocations from UI and let
us know the results. If issue persists pl. get the details of the error
customer is facing when trying to backout and we can review accordingly.

1. This message was sent from Oracle Development to Oracle Support.  Oracle Support then copied/pasted the message to pass it on to us.
2. LPN Context 5 is used on all LPNs that were used in cartonization.  We have been live for 1.5 years and have been doing this without a problem.  It seems we (the customer) know about the software the Oracle does
3. Why are they abbreviating “please” to “Pl.”?  That seems lazy and unprofessional to me.
4. They are asking us to backorder and re-release the whole delivery, essentially giving up on the problem.
5. By giving up, we must put material back into the warehouse, and repick it.  Causing us to use our resources 3x normal usage to pick an order. Luckily this delivery is only one LPN/Pallet.
6. If you can’t tell, their internal system puts arbitrary new lines into the text. And to force a new line, they must put a “.” in so that the new line isn’t truncated.  How antiquated is there support system? They market themselves as being a technology company.

Example 2

This service request is about an item master flag for enabling/disabling cycle counts.

Oracle designed their software in a very strange way.  In order to create a cycle count, the user must create a group of items, or an ABC Class/Compile.  Then the user must create a cycle count header, and assign the ABC Class/Compile.  Assigning the compile generates a distinct list of items that can be counted in that cycle count header.

If a user goes into an item master, and un-checks the “cycle count enabled” check box, the item can still be counted in a cycle count, because the item wasn’t removed from the cycle count items list. Apparently that check box impacts the ABC Compile creation only.

For our business process, when we obsolete an item, we would prefer not to allow a user to cycle count that item back into inventory, which is why we un-check that “cycle count enabled” box. It’s seems absurd to me that the list of values for items when entering cycle count wouldn’t take this check box into consideration. I opened an SR to report what I believed was a bug.

It took a few weeks of debate whether this was a bug or it was working as designed. It also took some debate to eliminate their concern that it would be a performance impact if they check the “cycle count enabled” check box every time a user enters a count. I told them they were already in the table. The user enters an item number, not an item ID, so they would be in the item master table checking that; why not include 1 line of code in the SQL to remove items where the cycle count is disabled?  AND CYCLE_COUNT_ENABLED_FLAG = ‘Y’  would do the trick.  

Recently I got this update on the service request:

Hello Ed –
Update from development :
” Customer test case is very corner.. Normally when there are cycle count
entries no customers will not update the Item set up. So Initially, we
thought making code changes will effect the existing functionality and it may
cause performance issues. However later due to customer push we have
analyzed more and we understand that code changes will not effect cause any
issues.
.
Please let me know whether customer wants fix for this issue or not? As per
latest update from customer if they are not ready to test the fix and if they
are not ready to take this fix then there is no point in doing code changes
and creating patches. If customer doesn’t need any code changes I will be
closing this bug.”

1. I really have no idea what they are trying to say in the first paragraph. It seems they maybe agree that there would be no performance problem after all?
2. How could the Oracle Support representative just pass this poorly written thing on to the customer?
3. The fix they are referring to in the second paragraph is a package that they asked us to test. I’m sorry, but the customer is not a beta tester. Oracle should be testing this solution on their own internal instances, not ours.  I guess if they want to refund some of our support contract fees, we might be open to that.
4. Why are we telling the software provider how to code their own software, and then paying them for the software?
5. Again, you can see the period that forces a new line.

Prior posts about Oracle Support Quality

• https://blog.edhayes.us/2015/03/25/oracle-support-quality/
• https://blog.edhayes.us/2014/04/03/working-effectively-with-customers-on-oracle-service-requests/