Re: About inheritance - Mailing list pgsql-advocacy

From Merlin Moncure
Subject Re: About inheritance
Date
Msg-id 6EE64EF3AB31D5448D0007DD34EEB34101AEA8@Herge.rcsinc.local
Whole thread Raw
In response to About inheritance  (Diogo Biazus <diogob@gmail.com>)
List pgsql-advocacy
> Merlin,
>
> > I can't really comment on 'ripping out' strategies, but I don't
think
> > there would be many complaints of inheritance was relegated to a
> > footnote in the docs and the default GUC setting was defaulted to
> > 'false'.
>
>  Why?   What harm is inheritance doing where it is?   Per Elein, Joe
and
> Chris, people *are* using it.

Note that I don't recommend ripping it out...at the very worst case it
should be deprecated for one or more versions (and I don't think it
should even go that far, unless there is a suitable replacement for just
about every use).  I'm just trying to look at things from a newcomer to
the database.

Joe and elein's uses notwithstanding, inheritance has always bugged me a
little bit because it has a certain implied use that it does not
fulfill.  Since TI is prominently displayed in the tutorial, a lot of
people try it out only to run smack into its limitations.  IOW, my
issues are mainly with the documentation, and how it causes postgres to
be perceived by people trying out the database.

Merlin



pgsql-advocacy by date:

Previous
From: Paul Thomas
Date:
Subject: Re: Fujitsu in news
Next
From: "Merlin Moncure"
Date:
Subject: Re: Fujitsu in news