Re: [HACKERS] logical decoding doc

2014-03-19 Thread Robert Haas
On Tue, Mar 18, 2014 at 7:27 PM, Tatsuo Ishii wrote: Maybe this is to demonstrating peek ahead does not consume changes, but IMO this is a little bit confusing for readers and I think there's a room to enhance the second sql session comment for example: I didn't write that

[HACKERS] logical decoding doc

2014-03-18 Thread Tatsuo Ishii
It seems two exactly same sql sessions are repeated in logicaldecoding.sgml. Is this intended? postgres=# -- You can also peek ahead in the change stream without consuming changes postgres=# SELECT * FROM pg_logical_slot_peek_changes('regression_slot', NULL, NULL); location | xid |