Re: Composite types or composite keys?

On 18 Nov 2013, at 2:24 pm, Chris Travers <chris(dot)travers(at)gmail(dot)com> wrote:
>
> I haven't done work with this so I am not 100% sure but it seems to me based on other uses I have for table inheritance that it might work well for enforcing interfaces for natural joins. The one caveat I can imagine is that there are two issues that occur to me there.
>
> 1. If you have two child tables which add a column of the same name, then your centralized enforcement gets messed up and you have a magic join which could take a while to debug....
>
> 2. The same goes if you have two child tables which also inherit a different parent table for a different natural join....
>
> To be honest I think being explicit about joins is usually a very good thing.

I can see how debugging a magic join would quickly outweigh any benefits and the “USING()” clause nicely reflects the foreign key definition, so I’ll stick with explicit joins.

Thanks,

Tony