Re: Thoughts about updateable views

Tom Lane wrote:
> Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us> writes:
> > I was thinking we create an updatable view if possible, and throw an
> > error if they try to insert/update/delete into a non-updatable view.
>
> Right.
>
> > Right now we ignore such activity, but that seems wrong.
>
> Nonsense.
>
> regression=# create view bar as select * from foo;
> CREATE VIEW
> regression=# insert into bar values (1);
> ERROR: cannot insert into a view
> HINT: You need an unconditional ON INSERT DO INSTEAD rule.
> regression=#
>
> The only backwards-compatibility risk I see is if there's a manually
> created ON-whatever rule; this probably has to override the automatic ones.

Oh, good. What release fixed this? I sure hope it is 7.5 because I
don't see any mention of it in the release notes. :-) I poked around
in CVS and can't see when it was added. No matter --- it is fixed now.

--
Bruce Momjian | http://candle.pha.pa.us
pgman(at)candle(dot)pha(dot)pa(dot)us | (610) 359-1001
+ If your life is a hard drive, | 13 Roberts Road
+ Christ can be your backup. | Newtown Square, Pennsylvania 19073