BUG #1409: A good and a bad news: Crazy SQL JOIN?
От | Lutischán Ferenc |
---|---|
Тема | BUG #1409: A good and a bad news: Crazy SQL JOIN? |
Дата | |
Msg-id | 200501180914.j0I9ENIU025570@developer.pgadmin.org обсуждение исходный текст |
Ответы |
Re: BUG #1409: A good and a bad news: Crazy SQL JOIN?
|
Список | pgsql-bugs |
The following bug has been logged online: Bug reference: 1409 Logged by: Lutischán Ferenc Email address: yoursoft@freemail.hu PostgreSQL version: 8.0 B5, 7.4.6 Operating system: WinXp, SLES9 Description: A good and a bad news: Crazy SQL JOIN? Details: Dear Developer Team! A good news: Refering to my mail with 'Out of memory problem' (>>If you make 'create user' and 'alter group' sql command on existing users, and make it many times, the server doesn't release the memory<<) -> this is solved in the 7.4.6. A bad news: I don't understand the following situation: Try to create these tables: ------------------------------------------- SET client_encoding = 'UNICODE'; SET check_function_bodies = false; SET client_min_messages = warning; CREATE SCHEMA ifc AUTHORIZATION postgres; SET search_path = ifc, pg_catalog; SET default_tablespace = ''; SET default_with_oids = true; CREATE TABLE test ( col1 character varying(10), col2 character varying(10) ); ALTER TABLE ifc.test OWNER TO postgres; CREATE TABLE test2 ( col1 character varying(10), col2 character varying(10) ); ALTER TABLE ifc.test2 OWNER TO postgres; COPY test (col1, col2) FROM stdin; b ac ba a \N aac \N aab \. COPY test2 (col1, col2) FROM stdin; b ac \N aac ba a \N aaa \. ----------------------------------------------- And try to make the following selects: ----------------------------------------- select a.col2 as col1, b.col2 from ifc.test a full outer join ifc.test2 b on a.col2=b.col2 order by b.col2 ----------------------------------------- select a.col2 as col1, b.col2 from ifc.test a full outer join ifc.test2 b on a.col2=b.col2 order by b.col2::char(8) ----------------------------------------- A questions: Why different are the results? Best Regards: Ferenc
В списке pgsql-bugs по дате отправления: