On Tue, Mar 10, 2015 at 5:53 PM, Jim Nasby <Jim.Nasby@bluetreble.com> wrote: > I don't think we need both array_offset and array_offset_start; can't both > SQL functions just call one C function?
Not if you want the opr_sanity tests to pass.
(But I'm seriously starting to wonder if that's actually a smart rule for us to be enforcing. It seems to be something of a pain in the neck, and I'm unclear as to whether it is preventing any real problem.)
It is simple protection against some oversights. I am not against this check - this rule cleans a interface between C and SQL. More, the additional C code is usually very short and trivial.
