Received: from sss.sss.pgh.pa.us (sss.pgh.pa.us [209.114.166.2]) by hub.org (8.9.3/8.9.3) with ESMTP id LAA54786 for ; Thu, 5 Aug 1999 11:48:08 -0400 (EDT) (envelope-from tgl@sss.pgh.pa.us) Received: from sss.sss.pgh.pa.us (localhost [127.0.0.1]) by sss.sss.pgh.pa.us (8.9.1/8.9.1) with ESMTP id LAA04649; Thu, 5 Aug 1999 11:47:02 -0400 (EDT) To: rob caSSon cc: pgsql-sql@postgresql.org Subject: Re: [SQL] primary key view failure In-reply-to: Your message of Thu, 5 Aug 1999 11:24:25 -0400 (EDT) Date: Thu, 05 Aug 1999 11:47:02 -0400 Message-ID: <4647.933868022@sss.pgh.pa.us> From: Tom Lane rob caSSon writes: > should have sent this along in the first place.... > using latest unstable debian package...6.5.1-4 OK, then it evidently is a bug or porting problem. Hard to say what though; we need more evidence. Is there any difference in the plans you get from "explain" of the direct query and the view query? I get this either way: Hash Join (cost=83.94 rows=19044 width=94) -> Nested Loop (cost=53.62 rows=552 width=58) -> Seq Scan on pg_index (cost=2.19 rows=1 width=20) -> Seq Scan on pg_attribute (cost=51.43 rows=1104 width=38) -> Hash (cost=6.55 rows=138 width=36) -> Seq Scan on pg_class (cost=6.55 rows=138 width=36) regards, tom lane