On 10/02/2026 12:52, Marcos Pegoraro wrote:
> I think the approach will fail at some point if you keep using CREATE
> TABLE LIKE, because so many things will be incomplete or at least
> strange, that you'll have to redo many of them.
> Sequencevalues will be shared with the old table.
> Constraints/Indexes names will have to be renamed.
> Partitioned tables will have to be attached.
> Foreign Keys are not created either.
Quite right. Either we draw a clear line here, or it will be a tough nut
to crack.
Best, Jim
