> > > As you had it, any nonzero result would prevent backtick substitution. > I'm not really sure how much thought went into the existing behavior, > but I am pretty sure that it's not part of this patch's charter to > change that. > > regards, tom lane >
The changes all make sense, thanks!
