Re: [HACKERS] Enhancing PGSQL to be compatible with Informix SQL

Поиск
Список
Период
Сортировка
От Bruce Momjian
Тема Re: [HACKERS] Enhancing PGSQL to be compatible with Informix SQL
Дата
Msg-id 200001062204.RAA22918@candle.pha.pa.us
обсуждение исходный текст
Ответ на RE: [HACKERS] Enhancing PGSQL to be compatible with Informix SQL  ("Ansley, Michael" <Michael.Ansley@intec.co.za>)
Список pgsql-hackers
> SELECT blah, blah, blah
> FROM t1, t2, t3, t4
> WHERE t1.start_date BETWEEN t2.start_date (+) AND t2.end_date (+)
> AND t1.y = t2.y (+)
> AND t3.x (+) = t1.x
> AND t3.y (+) = t1.y
> AND t4.x = t1.x;
> 
> For example...
> 
> I realise that this is not standard, but it's easy to read, and easy to
> develop.
> 
> The problem with OUTER is: OUTER on which relationship?  Does this matter?
> I haven't thought about it hugely, but it may not make sense when you try to
> do this:
> 
> SELECT * 
> FROM t1, OUTER t2, t3
> WHERE t1.x = t2.x
> AND t2.y = t3.y
> 
> Which is the OUTER join?  Outer joining to t1 and inner joining to t3 gives
> (I think) a different result to inner joining to t1 and outer joining to t3.
> Then you have to start creating language rules to help determine which join
> becomes the outer join, and it becomes a bit of a mess.  With Oracle's
> notation, it's pretty clear (I think anyway).

This must be why the ANSI standard requires you to specify the join when
doing outer.  Thomas says we are going only with ANSI syntax, and I can
see now why OUTER is just looking for problems.

--  Bruce Momjian                        |  http://www.op.net/~candle maillist@candle.pha.pa.us            |  (610)
853-3000+  If your life is a hard drive,     |  830 Blythe Avenue +  Christ can be your backup.        |  Drexel Hill,
Pennsylvania19026
 


В списке pgsql-hackers по дате отправления:

Предыдущее
От: The Hermit Hacker
Дата:
Сообщение: Re: [GENERAL] New Search Engine ... UdmSearch
Следующее
От: "Patrick Welche,SCC,ext.35710,"
Дата:
Сообщение: pg_dump problem