Re: SQL options precedence for SQLRPGLE

On 19 Jun 2013 09:01, Josh Diggs wrote:

<<SNIP>> What makes sense to me is that QAQQINI would provide
defaults, the compile command could override those defaults, and
the SET OPTION would be the final word. Is that correct? <<SNIP>>

What are the compile-time options that are in question; i.e. what are the alluded compile-time options that are coming from the QAQQINI and corresponding to an equivalent specification that would be established by either the SQL statement SET OPTION and\or [pre-]compiler parameter?

The Query Options feature is [mostly] for run-time; i.e. defines mostly, the effects of the Query Engine decision-making. So while an Access Plan may be built during compile-time while some Query Option is in effect, whatever are the Query Options that are or are not in effect when that Plan is rebuilt during run-time, are the new rules for the Query Engine; except possibly, any rules that are compile-time-only.

The Query Options exist to define *exceptions* to the default behavior. Thus they are not resolved for the default behavior. Instead they are resolved for overridden behavior. For consistency, the final arbiter of any run-time option is from the Query Options feature... which would mean that any option conflicting with a [or defining the same] feature, either as an SQL OPTION or [pre-]compiler parameter, would be overridden by the Query Option.