Okay, now it's more clear where you're coming from. It's just so common to hear of cases where applications that use RLA differ from alternatives that use SQL, that "of course" was the first thing that came to mind. But those cases are resolved by in depth analysis of the alternative applications. In your case, it appears that the RLA is not honoring the where clause of an SQL CREATE View. I've never tried limiting an SQL View to a sub-selection of records. It's actually surprising to hear that works.
-Nathan.
----- Original Message ----
From: James H. H. Lampert <jamesl@xxxxxxxxxxxxxxxxx>
To: Midrange Systems Technical Discussion <midrange-l@xxxxxxxxxxxx>
Sent: Wed, April 14, 2010 11:35:25 AM
Subject: Re: More fun with native RLA on an SQL view
Nathan Andelin wrote:
Of course.
Of course? That doesn't say much, and it's certainly contrary to any
reasonable expectation.
I'll be more specific:
The CREATE VIEW is of the form:
Create View &LIB/FROBOZZV2 as
Select P_Foo, P_Bar, P_Baz, P_Qux, P_Corge,
P_InActive, P_Grault, P_Garply, P_Plugh
From FROBOZZ
Where P_InActive = 'Y'
RcdFmt FROBOZZ;
(Copied verbatim from what the end-user sent me, even to the number of
fields in the view; "only the names have been changed, to protect the
innocent.")
And looking at the resulting view through native RLA, the results are as
expected except that there's one record with P_InActive = 'N'
--
JHHL
midrange.com
