I don't think there is any question about implementing ANSI syntax for outer
joins. There are two compelling reasons for this:
a) it is a stated goal of the PostgreSQL project to be ANSI compliant, and
to be a testbed for new ANSI features (this implies adherence to ANSI before
you can go extending it with new features)
b) it seems to be the most accurate (if not the prettiest) way of specifying
an outer join
For these reasons, ANSI should be used for the implementation. If, however,
we manage to provide a second, short-cut method for getting the same result,
so be it. However, this then begs the question: which short method to use.
Well, I propose that once outer joins have been implemented (using ANSI
syntax, many thanks to Bruce and Thomas) we have a quick vote to see what
people like, and whoever feels like implementing it can go for it. I would
also suggest that only one short method is used, because otherwise we'll
land up having to support n different syntaxes, each of which is not widely
enough used to justify the work.
MikeA
-----Original Message-----
From: Bruce Momjian
To: Rod Chamberlin
Cc: Thomas Lockhart; Don Baccus; Ansley, Michael; 'The Hermit Hacker ';
'pgsql-hackers@postgreSQL.org '
Sent: 1/7/00 6:23 PM
Subject: Re: [HACKERS] SQL outer join syntax
> > > I can't imagine how I would answer a question: "How do I do an
ANSI
> > > outer join". It would need its own FAQ page.
> >
> > Well, *you're* the one writing the book :))
> >
>
> I'd have thought this gave him justtification to complain about your
> horrible syntax then:)
The big problem is that is no Thomas's syntax, but the ANSI syntax, and
there doesn't seem to be any vendor-neutral solution for outer joins
other than the ANSI one.
-- 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,
Pennsylvania
19026