Re: Dropping a schema - Mailing list pgsql-hackers

From Tom Lane
Subject Re: Dropping a schema
Date
Msg-id 24343.1029946351@sss.pgh.pa.us
Whole thread Raw
In response to Re: Dropping a schema  (Oliver Elphick <olly@lfix.co.uk>)
List pgsql-hackers
Oliver Elphick <olly@lfix.co.uk> writes:
> On Wed, 2002-08-21 at 15:02, Tom Lane wrote:
>> Also, it seems the safest behavior to me.  "rmdir dir" won't remove a
>> nonempty directory; isn't that a pretty close analogy?

> Not really, seeing that you can't say "mkdir directory (containing these
> files)".  An implicit cascade *inside* the schema seems an appropriate
> parallel to "CREATE SCHEMA ... (CREATE TABLE ...)".  After all, we don't
> have to say "DROP TABLE ... CASCADE" because the table has rows in it!

Hm.  I could see an argument for being willing to auto-drop stuff that
had been made that way (inside CREATE SCHEMA) but not stuff that was
made by separate commands.  But the spec doesn't seem to make any such
distinction: RESTRICT is RESTRICT.  In any case, I like the behavior as
it is, so I'm not gonna go out of my way to change it...
        regards, tom lane


pgsql-hackers by date:

Previous
From: "Zeugswetter Andreas SB SD"
Date:
Subject: Re: @(#) Mordred Labs advisory 0x0001: Buffer overflow in
Next
From: Justin Clift
Date:
Subject: Re: @(#)Mordred Labs advisory 0x0003: Buffer overflow in