Re: [NOVICE] Last ID Problem - Mailing list pgsql-hackers

From Tom Lane
Subject Re: [NOVICE] Last ID Problem
Date
Msg-id 6531.1107378267@sss.pgh.pa.us
Whole thread Raw
In response to Re: [NOVICE] Last ID Problem  (Greg Stark <gsstark@mit.edu>)
Responses Re: [NOVICE] Last ID Problem  ("Greg Sabino Mullane" <greg@turnstep.com>)
List pgsql-hackers
Greg Stark <gsstark@mit.edu> writes:
> This is from the DBI documentation -- that is, the non-driver-specific
> abstract interface documentation.

>            Returns a value 'identifying' the row just inserted, if possible.
>            Typically this would be a value assigned by the database server to
>            a column with an auto_increment or serial type.

Aside from the numerous serious problems pointed out in the
documentation, this has an even more fatal objection, which is that it's
unspecified what the result value is and thus there is no portable way
of *using* the result after you have it.  (If the PG driver returns an
OID you certainly couldn't use that the same way as some other driver
that returns a primary key ... especially a multicolumn primary key ...)

This "portable" function is so unportable that I see no reason to
accept it as precedent.
        regards, tom lane


pgsql-hackers by date:

Previous
From: Oliver Jowett
Date:
Subject: Re: [NOVICE] Last ID Problem
Next
From: Tom Lane
Date:
Subject: Re: pg_dump bug in 7.3.9 with sequences