Robert Haas <robertmhaas@gmail.com> writes:
> I think that the issue here is simply that because both the updated
> table and the "excluded" pseudo-table are visible here, and have the
> same columns, an unqualified name is ambiguous. I don't really think
> that it's worth documenting. The error message you get if you fail to
> do it is actually pretty good:
> ERROR: column reference "b" is ambiguous
> Now you could read that and not understand that the ambiguity is
> between the target table and the "excluded" pseudo-table, for sure.
Agreed. It doesn't help that there's no explicit use of "excluded"
anywhere, as there is in more usual ambiguous-column cases.
> What would probably help more is adding something like this to the
> error message:
> HINT: column "b" could refer to any of these relations: "foo", "excluded"
+1, that seems like it could be handy across the board.
regards, tom lane