On Wed, Aug 22, 2012 at 10:22 PM, Chris Travers <chris.travers@gmail.com> wrote:
> I have now been working with table inheritance for a while and after
> starting to grapple with many of the use cases it has have become
> increasingly impressed with this feature. I also think that some of
> the apparent limitations fundamentally follow from the support for
> multiple inheritance, and multiple inheritance itself is so useful I
> would not want to see this go away. Inheritance really starts to come
> to its own once you start using table methods, and some features that
> are useful in some sorts of inheritance modelling are useless in
> others.
The problem with postgres table inheritance is that it doesn't really
solve the problem that people wanted solved: to be able to define an
set of specific extra attributes for each row depending on some
characteristic of that row. The feature only tantalizingly
masquerades as such.
Until it found use in table partitioning, I found the inheritance
feature to be basically useless.
merlin
