"Merlin Moncure" <mmoncure@gmail.com> writes:
> On Jan 27, 2008 10:45 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>> If we were to change this, we'd probably have to think in terms of
>> making the active search_path be part of the lookup key for cached plans.
> For the record, IMO it would on balance be better to have the plan
> invalidate when setting the search path.
I think that the actual use-case for this would likely involve
repetitive execution of a function F against various search_path
settings, and so what we'd want is to cache the appropriate plan for
each path setting, not just blow away the whole cache when search_path
changes. But the whole thing is something to investigate for 8.4.
regards, tom lane