Ross, you make some powerful arguments here. Probably the most
significant was the idea that you need a unique identifier for every
row, and it should be of a consistent type, which primary key is not.
We clearly need a GUC parameter to turn on/off oids. But it seems we
will always need the abi
On Thu, Jan 23, 2003 at 10:03:28AM -0500, Tom Lane wrote:
> Curt Sampson <[EMAIL PROTECTED]> writes:
> > I object. I personally think we should be moving towards not using OIDs
> > as the default behaviour, inasmuch as we can, for several reasons:
>
> All these objections are global in nature, not
Curt Sampson <[EMAIL PROTECTED]> writes:
> I object. I personally think we should be moving towards not using OIDs
> as the default behaviour, inasmuch as we can, for several reasons:
All these objections are global in nature, not specific to CREATE TABLE
AS. The argument that persuaded me to do
> "Christopher Kings-Lynne" <[EMAIL PROTECTED]> writes:
> > Why don't you just include them by default, otherwise if WITHOUT OIDS
> > appears in the CREATE TABLE command, then don't include them ?
>
> Well, adding a WITHOUT OIDS option to CREATE TABLE AS would be a new
> feature, which I don't have
"Christopher Kings-Lynne" <[EMAIL PROTECTED]> writes:
> Why don't you just include them by default, otherwise if WITHOUT OIDS
> appears in the CREATE TABLE command, then don't include them ?
Well, adding a WITHOUT OIDS option to CREATE TABLE AS would be a new
feature, which I don't have the time/i
Why don't you just include them by default, otherwise if WITHOUT OIDS
appears in the CREATE TABLE command, then don't include them ?
Chris
> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of Tom Lane
> Sent: Wednesday, 22 January 2003 4:12 AM
> To: [EMA
