Re: BUG #16368: Incorrect function inlining in the presence of a window function

> The particular case mentioned there seems to be about the same as
> here: constant-folding happens even in CASE arms that will never be
> reached at runtime.  

Since the actual issue here is false positive exceptions, why can't we 
treat errors raised during const eval as a signal to cancel the constant 
folding, i.e. ignore them and keep the failing expression unoptimized?

                              Elvis