On 9/30/16 4:32 PM, Thomas Munro wrote:
>> Hmm, yeah, that will need more work. To be completely correct, I think,
>> > we'd also need to record the version in each expression node, so that
>> > check constraints of the form CHECK (x > 'abc') can be handled.
> Hmm. That is quite a rabbit hole. In theory you need to recheck such
> a constraint, but it's not at all clear when you should recheck and
> what you should do about it if it fails. Similar for the future
> PARTITION feature.
I think it's not worth dealing with this in that much detail at the
moment. It's not like the collation will just randomly change under you
(unlike with glibc). It would have to involve pg_upgrade, physical
replication, or a rebuilt installation. So I think I will change the
message to something to the effect of "however you got here, you can't
do that". We can develop some recipes and ideas on the side for how to
recover situations like that and then maybe integrate tooling for that
later.
--
Peter Eisentraut http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services