Re: BUG #3048: pg_dump dumps intarray metadata incorrectly

Oleg Bartunov wrote:
> On Mon, 2 Apr 2007, Bruce Momjian wrote:
>
> >Tom Lane wrote:
> >>Bruce Momjian <bruce(at)momjian(dot)us> writes:
> >>>Tom, do you want this fixed for 8.2.X?
> >>
> >>>Tom Lane wrote:
> >>>>Yeah. I'd say that intarray's attempt to override the default status of
> >>>>the built-in gin opclass is simply a bad idea and should be removed.
> >>
> >>I don't recall having seen a response from Oleg or Teodor, and would
> >>like their input before making a final decision --- but at the moment
> >>I think we should take that out.
> >
> >Oleg or Teodor, I need a comment on this.
>
> We agree with Tom in this case and we'll remove update of
> system catalog. But, I want to rise the problem again - pg_dump doesn't
> track changes of system catalog. The problem could be more pronounced in
> case of built-in FTS, if somebody with superuser rights changes
> fts configurations in system catalog.

But it should dump them as "ALTER whatever" commands, right?

--
Alvaro Herrera http://www.CommandPrompt.com/
The PostgreSQL Company - Command Prompt, Inc.