Number of problems by platform (was: DB2/400 comparisons with other relational databases (specifically Oracle and SQL Server))

>> 1)  I don't take grief.  I see a problem - I report it.
>> 2)  Maybe we push the envelope a little harder.  More data, different
>> applications, more sql.
>
>I don't take grief, either.  I just don't seem to run into the
>problems you do.  And I was the Manager of Architecture for the
>world's largest AS/400 software company - we pushed the
>envelope, too.  BPCS, AS/SET and AS/NET definitely used some
>of the more advanced features of the AS/400, but that was a while ago.

I spent 17 years with a local wholesale paper distributor.  We opened
several APARs over those years (multiple midrange platforms.)  Now that I
work for a software developer, I have never had to open an APAR _here_.  I
have had customers open APARS because something that worked fine here did
not work out there.

This incredibly sophisticated analysis <grin> would seem to indicate that
the larger the operation is, the more chances there are to run into an
unexpected situation.  The developers simply don't have that large and
diverse a database that gets hammered 24x7.  My experience with customer
support bears this out as well.  Our small clients rarely report problems
with our software and almost never with IBM software.  The big ones manage
to find something with every release; some bugs are years old but haven't
surfaced until the right combination of drivers, data and storage conditions
were met.  It's the big customers who find the PTFable situations, too.

Some of this is because the client is back-level on OS/400 or on PTFs, but
some of it is because it's an IBM bug that takes precise circumstances to
materialise.

My thought for the month.  Good thing, too - I was afraid I wasn't going to
have one at all this month!
  --buck