On 11/13/17 02:33, Amit Langote wrote:
> On 2017/11/11 18:11, Michael Paquier wrote:
>> On Fri, Nov 10, 2017 at 11:33 AM, Michael Paquier
>> <michael.paquier@gmail.com> wrote:
>>> All in all, I would be more on board with just issuing errors when
>>> trying to attempt typed table creation with identity columns and
>>> partitions for now, by complaining with a
>>> ERRCODE_FEATURE_NOT_SUPPORTED. Any person willing to add this support
>>> later on could always provide a patch for it, I am now rather
>>> convinced that what I proposed upthread is not the correct way to do
>>> so, and there is no rush in implementing it correctly for v11 or even
>>> later.
>>>
>>> Peter, Robert, Amit, thoughts?
>>
>> And this gives the attached. That still looks like the safest thing to
>> do for now on v10 and HEAD after more thoughts.
>
> That looks good to me. That is, if the consensus turns out to be that we
> issue error in those cases.
committed
--
Peter Eisentraut http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
