> Of course, this is made much easier if there is a pg_depends table
> which accurately records dependencies.
Yes, that was a nifty idea.
> I have a meta-point: the choices to be made here are not all that
> interesting. They do have to be defined. But almost any definition
> is OK. Users are not going to routinely redefine tables with attached
> functions; when they do, they must be prepared to consider the
> consequences. If anybody thinks that different choices should be made
> in this case, that is certainly fine with me.
>
> If you agree with me on the meta-point, then this is just a quibble
> about my original patch (which made choice 1 above). If you disagree
> with me, I'd like to understand why.
I agree that having problems when a table is defined is acceptable. It
is not like someone is _forced_ to use the feature.
So far that is three or four people who like the feature, and I have
only heard one opposed.
-- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610)
853-3000+ If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill,
Pennsylvania19026