On Wed, Mar 7, 2012 at 2:49 PM, Merlin Moncure <mmoncure@gmail.com> wrote:
> On a practical level, the error blocks nothing -- you can bypass it
> trivially. It's just an annoyance that prevents things that users
> would like to be able to do with table row types. So I'd argue to
> remove the check, although I can kinda see the argument that it's not
> a bug unless the check was recently introduced so that it broke older
> code.
The behavior hasn't changed since at least as far back as 8.1, so
you're correct (once again) -- not a bug. I'm really surprised I
haven't already bumped into this. I usually don't mix
tables-as-storage with tables-as-composites though.
Mike, on 9.1, you'll probably get more mileage out of using the hstore
type for row storage if you want to do auditing in that style.
merlin