Jim Nasby wrote:
> FWIW, my interest in this is largely due to the problems I've had with this
> in the variant type. In particular, using the same resolution rules for
> functions and operators. So I'm wondering if there's a bigger issue here.
I'd be glad to review your variant stuff, but I doubt it's the same
thing at play. Please don't hijack this thread. Deparse is doing some
pretty nasty games with the backend innards to make the deparsed
representation usable generally. We were just missing a simple trick.
--
Álvaro Herrera http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
