On 2/24/2017 11:41 AM, Steinmetz, Paul wrote:
It was mentioned about storing source as DDL instead of DDS.
Anyone doing this?
Yes, ish.
We are not not not I say again not converting DDS to DDL. New tables,
for new applications, with new designs are being done in DDL, and the
application code is written to take advantage of the benefits of
SQL/DB2. Here endeth the sermon.
How is done?
IFS directory tree, with a hierarchy that roughly matches the
applications. So one for A/R, one for customer care, etc. We map the
base of this tree to a Windows drive letter and use the ACS SQL script
tool to execute it. This also allows us to use RDi on the files as
well. Actually, any PC based editor can work on these. EDTF/DSPF works
from the green screen side.
The files themselves have a .SQL extension. Originally, we stored DDL
in a file named after the DB2 'thing' we're creating, so
customer.sql
customer_sales_view.sql
getCustomer_procedure.sql
This tends to generate a lot of tiny 'source' files. If the naming
convention is... iffy, I use RDi's Remote Search to search through /all/
of them to find the file containing the DDL I'm interested in.
For the most part, I don't ever do a DROP TABLE...CREATE TABLE. When I
need to add columns to an existing table, I do an ALTER TABLE. I do
CREATE OR REPLACE for views, indexes, functions, and procedures. I think
that hits the high points.
midrange.com
