On 25-Aug-2014 15:51 -0500, Bill Howie wrote:
In the SQL scenario, how would the actual copying out of the data be
accomplished? Via a CREATE TABLE within the SQL scripting somewhere?
If running with isolation in order to enable a DROP TABLE to be
effected with COMMIT after determining there was no data, then probably
simplest with CREATE TABLE ... AS /* or LIKE */ ... request.
However, actual requirements may dictate otherwise; e.g. for /exact/
matching to DDS-created ERP PFs such that Create Duplicate Object
(CRTDUPOBJ) might be preferable. But there are many more considerations
than just DDS PF vs SQL TABLE, which even the parameters on CRTDUPOBJ
allude, such as for triggers and constraints; as well there are
copy-options for the CREATE TABLE that should be considered. Difficult
for the casual observer to guess the requirements, such that anything
perceived "more elegant" might not even be an option.