Neil Conway wrote:
> On Fri, 2003-01-10 at 13:32, Bruce Momjian wrote:
> > I can't think of any logic that would allow a table to be skipped on a
> > dump, same with namespaces.
>
> Uh, what? The issue is that:
>
> pg_dump -t some_non_existent_table_name
>
> produces an empty dump (well, it has a 3 lines of comments, but no SQL).
> Similarly:
>
> pg_dump -n some_non_existent_schema_name
>
> also produces an empty dump. I'm asking if that's the correct behavior:
> perhaps we should throw an error instead?
We certainly should throw an error!
--
Bruce Momjian | http://candle.pha.pa.us
pgman@candle.pha.pa.us | (610) 359-1001
+ If your life is a hard drive, | 13 Roberts Road
+ Christ can be your backup. | Newtown Square, Pennsylvania 19073
