> The original naming complaint reflected a concern that
> the symbol looked like it was supplied by the system headers, rather
> than being of Postgres origin. Heikki's suggestion deals with that,
> and I think it's fine as-is.
OK, fine with me.
--Ian