Re: Temparary disable constraint - Mailing list pgsql-hackers

From Adnan DURSUN
Subject Re: Temparary disable constraint
Date
Msg-id BAY106-DAV119B8C02FE8CD82F7D0573FAAA0@phx.gbl
Whole thread Raw
In response to Re: Temparary disable constraint  (Bruce Momjian <bruce@momjian.us>)
List pgsql-hackers
----- Original Message ----- 
From: "Richard Huxton" <dev@archonet.com>
To: "Bruce Momjian" <bruce@momjian.us>
Cc: "Adnan DURSUN" <a_dursun@hotmail.com>; <pgsql-hackers@postgresql.org>
Sent: Thursday, January 18, 2007 5:57 PM
Subject: Re: [HACKERS] Temparary disable constraint


>> Not easily, because the view are bound to the object id of the tables
>> involved.
>
> The trick would be I think to bind them to individual columns, so if view 
> V doesn't mention column C then dropping C has no effect on it.
>
> That's a lot more dependencies to track of course.
>  Is that not possible that all objects have a column that says whether 
object state is ok or not.
When any session wants to try to execute any DML on object, then the 
Postgres checks that state column.
If object state is not ok then Postgres raises an exception.

Best Regards

Adnan DURSUN
ASRIN Bilisim Ltd. 



pgsql-hackers by date:

Previous
From: Teodor Sigaev
Date:
Subject: Re: Design notes for EquivalenceClasses
Next
From: Tatsuo Ishii
Date:
Subject: Re: [COMMITTERS] pgsql: Fix failure due to accessing an