"Zeugswetter Andreas DAZ SD" <ZeugswetterA@spardat.at> writes:
>> How are you planning to represent the association between MIN/MAX and
>> particular index orderings in the system catalogs?
> Don't we already have that info to decide whether an index handles
> an "ORDER BY" without a sort node ?
We know how to determine that an index matches an ORDER BY clause.
But what has an aggregate called MAX() got to do with ORDER BY? Magic
assumptions about operators named "<" are not acceptable answers; there
has to be a traceable connection in the catalogs.
As a real-world example of why I won't hold still for hard-wiring this:
a complex-number data type might have btree opclasses allowing it to be
sorted either by real part or by absolute value. One might then define
max_real() and max_abs() aggregates on the type. It should be possible
to optimize such aggregates the same way as any other max() aggregate.
regards, tom lane
