Re: [HACKERS] Confusing remark about UPSERT in fdwhandler.sgml

2015-10-04 Thread Etsuro Fujita
On 2015/10/03 5:57, Robert Haas wrote: On Fri, Oct 2, 2015 at 4:04 AM, Peter Geoghegan wrote: On Fri, Oct 2, 2015 at 1:00 AM, Etsuro Fujita wrote: ISTM that the sentence "as remote constraints are not locally known" is somewhat confusing, because

Re: [HACKERS] Confusing remark about UPSERT in fdwhandler.sgml

2015-10-02 Thread Robert Haas
On Fri, Oct 2, 2015 at 4:04 AM, Peter Geoghegan wrote: > On Fri, Oct 2, 2015 at 1:00 AM, Etsuro Fujita > wrote: >> ISTM that the sentence "as remote constraints are not locally known" is >> somewhat confusing, because check constrains on remote

[HACKERS] Confusing remark about UPSERT in fdwhandler.sgml

2015-10-02 Thread Etsuro Fujita
The following is a remark about UPSERT in fdwhandler.sgml. INSERT with an ON CONFLICT clause does not support specifying the conflict target, as remote constraints are not locally known. This in turn implies that ON CONFLICT DO UPDATE is not supported, since the specification

Re: [HACKERS] Confusing remark about UPSERT in fdwhandler.sgml

2015-10-02 Thread Peter Geoghegan
On Fri, Oct 2, 2015 at 1:00 AM, Etsuro Fujita wrote: > ISTM that the sentence "as remote constraints are not locally known" is > somewhat confusing, because check constrains on remote tables can be > defined locally in 9.5. How about "unique constraints or exclusion