Received: from vpcit.ru (qmailr@dialup.vpcit.ru [195.38.52.246]) by hub.org (8.9.3/8.9.3) with SMTP id FAA58072 for ; Thu, 3 Feb 2000 05:04:37 -0500 (EST) (envelope-from yura@vpcit.ru) Received: (qmail 31886 invoked from network); 3 Feb 2000 10:04:34 -0000 Received: from gercon.vpcit.ru (195.38.52.226) by dialup.vpcit.ru with SMTP; 3 Feb 2000 10:04:34 -0000 Date: Thu, 3 Feb 2000 15:03:57 +0500 From: Yury Don X-Mailer: The Bat! (v1.38e) S/N A1D26E39 / Educational Reply-To: Yury Don Organization: Gercon X-Priority: 3 (Normal) Message-ID: <3627.000203@vpcit.ru> To: pgsql-general Subject: Re: [GENERAL] Proposed Changes to PostgreSQL In-reply-To: <3898DA32.793A98BA@nimrod.itg.telecom.com.au> References: <3898DA32.793A98BA@nimrod.itg.telecom.com.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hello Chris, Once, Thursday, February 03, 2000, 6:30:26 AM, you wrote: CB> 1) An imaginary field in every tuple that tells you the class it came CB> from. CB> This is useful when you select from table* and want to know which CB> relation the object actually came from. It wouldn't be stored on disk, CB> and like oid it wouldn't be displayed when you do SELECT *. The field CB> would be called classname. So you could have... CB> SELECT p.classname, p.name FROM person p; CB> person | Fred CB> student | Bill CB> employee | Jim CB> person | Chris I am voting for this by both hands. Now we forced to use an additional column classname in every table and rule to fill this column. CB> 2) Changing the sense of the default for getting inherited tuples. CB> Currently you only get inherited tuples if you specify "tablename*". CB> This would be changed so that you get all sub-class tuples too by CB> default unless you specify "ONLY tablename". There are several CB> rationale for this. Firstly this is what Illustra/Informix have CB> implemented. Secondly, I believe it is more logical from an OO CB> perspective as well as giving a more useful default. If a politician CB> IS a person and I say SELECT * from person, then logically I should CB> see all the politicians because they are people too (so they claim CB> :). Thirdly, there are a whole range of SQL statements that should CB> probably be disallowed without including sub-classes. e.g. an ALTER CB> TABLE ADD COLUMN that does not include sub-classes is almost certainly CB> undesirable. It seems ashame to have to resort to non-standard SQL CB> with the "*" syntax in this case when it is really your only CB> choice. Basicly, wanting ONLY a classname is a far more unusual CB> choice, and leaving off the "*" is a common error. Fourthly, it seems CB> out of character for the SQL language to have this single character CB> operator. The SQL style is to use wordy descriptions of the operators CB> meaning. "ONLY" fits well here because it describes its own meaning CB> perfectly whereas to the unitiated, "*" is harder to guess at. While CB> this change is an incompatibility I hope for those few people using CB> inheritance they can accept the need to move forward without CB> over-burden of backwards compatibility. Sounds very logically. -- Best regards, Yury ICQ 11831432 mailto:yura@vpcit.ru