Thomas Swan <tswan@olemiss.edu> writes:
> The reason I was asking is that I had an idea for doing the select ** from
> tablename* that would expand.
> It could be macro of sorts but part of it depending on creating a null
> table or the equivalent of it with nothing but a null column for each
> different column of the set.
What happens when two different child tables have similarly-named
columns of different types?
In any case, this wouldn't be a very satisfactory solution because you
couldn't tell the difference between a null stored in a child table and
the lack of any column at all. We really need to do it the hard way,
ie, issue a new tuple descriptor as we pass into each new child table.
There appears to have once been support for that back in the Berkeley
days; you might care to dig through Postgres 4.2 or so to see how they
did it.
regards, tom lane