Thread: dlgTable_patch

dlgTable_patch

From
"Hiroshi Saito"
Date:
Hi Andreas.

It can't be operated by the table which has one column.
Operation is drop.
This becomes the table of the sky.
You should prevent Drop operation from doing it if it is presumed injustice
operation
though the intention of this operation is strange.

I think that it should do this patch from the flow.
Which is better?

Regards,
Hiroshi Saito

Attachment

Re: dlgTable_patch

From
Andreas Pflug
Date:
Hiroshi Saito wrote:

>Hi Andreas.
>
>It can't be operated by the table which has one column.
>Operation is drop.
>This becomes the table of the sky.
>You should prevent Drop operation from doing it if it is presumed injustice
>operation
>though the intention of this operation is strange.
>
>I think that it should do this patch from the flow.
>Which is better?
>
>
Hi Hiroshi,

I can't see a problem about the present code (with 7.3), a table must
have at least one column. If you dropped the last in that columns list,
you must add another to have a working table.
I just added the new 7.4's feature, being able to have zero columns
(whatever that's good for...).

Regards,
Andreas


Re: dlgTable_patch

From
"Dave Page"
Date:

> -----Original Message-----
> From: Andreas Pflug [mailto:pgadmin@pse-consulting.de]
> Sent: 08 August 2003 09:39
> To: Hiroshi Saito
> Cc: pgadmin-hackers@postgresql.org
> Subject: Re: [pgadmin-hackers] dlgTable_patch
>
>
> I can't see a problem about the present code (with 7.3), a table must
> have at least one column. If you dropped the last in that
> columns list,
> you must add another to have a working table.
> I just added the new 7.4's feature, being able to have zero columns
> (whatever that's good for...).

Multiple inherited tables, but no columns of it's own.

Regards, Dave.

Re: dlgTable_patch

From
Andreas Pflug
Date:
Dave Page wrote:

>>I just added the new 7.4's feature, being able to have zero columns
>>(whatever that's good for...).
>>
>>
>
>Multiple inherited tables, but no columns of it's own.
>
>
dlgTable will show inherited columns, so that table isn't considered empty.

PostgreSQL 7.4 now allows
CREATE TABLE empty()

Could be used as an anchor for rules, just as CREATE VIEW does.

Regards,
Andreas