Re: Mutable foreign key constraints

On 9/13/24 15:05, Andreas Karlsson wrote:
> On 9/13/24 4:41 AM, Laurenz Albe wrote:
>> That's very broken and should not be allowed.
>
> +1
>
>>> A possible objection is that if anybody has such a setup and
>>> hasn't noticed a problem because they never change their
>>> timezone setting, they might not appreciate us breaking it.
>>
>> I hope that there are few cases of that in the field, and I think it
>> is OK to break them.  After all, it can be fixed with a simple
>>
>>    ALTER TABLE foo ALTER id TYPE timestamptz;
>>
>> If the session time zone is UTC, that wouldn't even require a rewrite.
>>
>> I agree that it cannot be backpatched.
>
> I unfortunately suspect there might be more cases than we think in the
> field due to many people not understanding the difference between
> timestamp and timestamptz but the good thing is that
> timestamp/timestamptz are rare in foreign keys, even in composite ones.

It will become a lot more common with WITHOUT OVERLAPS, so I think it is
important to fix this at the same time or earlier as that feature.

> Since this is quite broken and does not have any real world usefulness I
> think we should just go ahead and disallow it and have a few people
> complain.

+1
--
Vik Fearing