Re: [SQL] Uniqueness of rule, constraint, and trigger names - Mailing list pgsql-hackers

From Dan Langille
Subject Re: [SQL] Uniqueness of rule, constraint, and trigger names
Date
Msg-id 200203042050.g24Kogk80096@lists.unixathome.org
Whole thread Raw
In response to Re: [SQL] Uniqueness of rule, constraint, and trigger names  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
On 4 Mar 2002 at 15:47, Tom Lane wrote:

> "Dan Langille" <dan@langille.org> writes:
> > It would be easily detected at load time.  Then you direct the user to
> > the tool mentioned above.
> 
> But if the user has already dumped and flushed his old installation,
> he's still in need of a time machine.

Agreed.  I had that same problem with upgrading to 7.2.  My solution was 
to install 7.1.3 on another box, load, massage, dump.  Next time, I'll 
keep the old version around a bit longer.

> Hmm, maybe what we need is a tool that can be applied during load.
> Essentially, alter incoming constraint names as needed to make them
> unique.  We could have this enabled by a SET switch...

I think that is an eloquent solution.
-- 
Dan Langille
The FreeBSD Diary - http://freebsddiary.org/ - practical examples



pgsql-hackers by date:

Previous
From: "Dan Langille"
Date:
Subject: Re: [SQL] Uniqueness of rule, constraint, and trigger names
Next
From: egon.phillips@sympatico.ca (egon.phillips)
Date:
Subject: Re: Problems with 7.2 on Solaris 2.6