Hi team. Thanks for the information.
Looks like there're some architectural limitations for such foreign keys.
Also thanks for the suggestions on how to make it behaving like I want on
current postgres version.
On Sat, 23 Nov 2019, 19:11 Tom Lane, wrote:
> Adrian Klaver writes:
> > Please reply
As a workaround, create a table with only one column and one value = `false`
and foreign to it.
On 22.11.2019 16:32, aleksey ksenzov wrote:
Latest time we faced several issues which wouldn't arise provided we have
possibility to use constants in foreign key constraints.
brief example where it
Adrian Klaver writes:
> Please reply to list also.
> Ccing list.
> On 11/22/19 11:36 PM, aleksey ksenzov wrote:
>> While I understand I can do everything with triggers/functions, for me
>> it looks like a good idea to have possibility to use constants in
>> constraints, so it would be very nice
On 11/22/19 11:36 PM, aleksey ksenzov wrote:
Please reply to list also.
Ccing list.
We're already on 12.
While I understand I can do everything with triggers/functions, for me
it looks like a good idea to have possibility to use constants in
constraints, so it would be very nice if postgres
On 11/22/19 6:32 AM, aleksey ksenzov wrote:
Hi team.
Latest time we faced several issues which wouldn't arise provided we
have possibility to use constants in foreign key constraints.
brief example where it would be helpful:
table_a
( id uuid,
parent_id uuid,
is_deleted boolean
)
having possib
Hi team.
Latest time we faced several issues which wouldn't arise provided we have
possibility to use constants in foreign key constraints.
brief example where it would be helpful:
table_a
( id uuid,
parent_id uuid,
is_deleted boolean
)
having possibility of FK (parent_id, false) to (id, is_delete
