On Thu, 30 Aug 2001, Jon Lapham wrote:
> Okay, thanks, dropping and recreating the rule worked.
>
> After thinking a bit about this, it would seem that the 'problem' is
> that I was *able* to drop a table that had rules referencing it. Would
> it be possible to either not allow this, or to issue some type of
> warning message? Otherwise, you go down the path of this (for me
> anyway) subtle problem.
The problem is right now we don't keep track of that sort of information
in any really usable way (apart from scanning all objects that might refer
to an oid). There've been discussions on -hackers in the past about this
and it should be on the todo list.
> Also, who should I send documentation patches to about this? I couldn't
> find any mention of this issue in the "create rule" documentation (or am
> I looking in the wrong place?) or in "Chapter 17: The Postgres Rule
> System". Hmmm, further perusal shows that Jan Weick is the author of
> the Chapter 17 documentation, I guess I can send text to Jan.
You might as well send patches to pgsql-patches and let everyone see them.