From: | Vik Fearing <vik(at)postgresfriends(dot)org> |
---|---|
To: | Corey Huinker <corey(dot)huinker(at)gmail(dot)com>, Trevor Gross <t(dot)gross35(at)gmail(dot)com> |
Cc: | PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org> |
Subject: | Re: WIP: System Versioned Temporal Table |
Date: | 2022-02-08 23:27:36 |
Message-ID: | 41d04d58-50ff-e794-93e8-ff59f35f646f@postgresfriends.org |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
On 1/24/22 00:16, Corey Huinker wrote:
>> - Table schemas change, and all (SV active) AV items would logically
>> need to fit the active schema or be updated to do so. Different story
>> for SV, nothing there should ever need to be changed.
>>
> Yeah, there's a mess (which you state below) about what happens if you
> create a table and then rename a column, or drop a column and add a
> same-named column back of another type at a later date, etc. In theory,
> this means that the valid set of columns and their types changes according
> to the time range specified. I may not be remembering correctly, but Vik
> stated that the SQL spec seemed to imply that you had to track all those
> things.
The spec does not allow schema changes at all on a a system versioned
table, except to change the system versioning itself.
--
Vik Fearing
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2022-02-08 23:42:33 | Re: [RFC] building postgres with meson - perl embedding |
Previous Message | Robert Haas | 2022-02-08 21:54:08 | Re: Fix BUG #17335: Duplicate result rows in Gather node |