Re: dblink_build_sql_update versus dropped columns

On 06/14/2010 11:21 AM, Tom Lane wrote:
> Actually, I was working on it myself. On further reflection I think
> that logical numbers are clearly the right thing --- if we define it
> as being physical numbers then we will have headaches in the future
> when/if we support rearranging columns. However, there is some small
> chance of breaking things in existing DBs if we back-patch that change.
> Thoughts?

I didn't even think people were using those functions for many years
since I never heard any complaints. I'd say better to not backpatch
changes to logical ordering, but FWIW the attached at least fixes the
immediate bug in head and ought to work at least a few branches.

> It strikes me also that the code is not nearly careful enough about
> defending itself against garbage input in the primary_key_attnums
> argument ...

Probably not :-(

Joe