From: | Guillaume Lelarge <guillaume(at)lelarge(dot)info> |
---|---|
To: | Thom Brown <thom(at)linux(dot)com> |
Cc: | Jan-Peter Seifert <Jan-Peter(dot)Seifert(at)gmx(dot)de>, pgadmin-support(at)postgresql(dot)org |
Subject: | Re: Feature request - Edit data and tables without a primary key |
Date: | 2012-07-14 13:09:57 |
Message-ID: | 1342271397.8088.20.camel@localhost.localdomain |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgadmin-support |
On Sat, 2012-07-14 at 14:01 +0100, Thom Brown wrote:
> On 14 July 2012 13:57, Guillaume Lelarge <guillaume(at)lelarge(dot)info> wrote:
> > On Thu, 2012-07-12 at 12:31 +0100, Thom Brown wrote:
> >> Use of the CTID system column would work.
> >
> > I don't see how that could work. CTIDs obviously are unique, but anyone,
> > updating a line will get another CTID for this line. If you try to
> > update it afterwards, in the best case, it errors out because there is
> > no line at this CTID. In the worst case, you update a completely
> > different line.
>
> Why can't you just use "RETURNING ctid" at the end of the statement to
> get the new one and update the grid with that value?
At the end of which statement? how do you want to make it work?
--
Guillaume
http://blog.guillaume.lelarge.info
http://www.dalibo.com
From | Date | Subject | |
---|---|---|---|
Next Message | Jan-Peter Seifert | 2012-07-14 13:43:03 | Re: Feature request - Edit data and tables without a primary key |
Previous Message | Guillaume Lelarge | 2012-07-14 13:09:07 | Re: Feature request - Edit data and tables without a primary key |