Received: from localhost (unknown [200.46.204.183]) by postgresql.org (Postfix) with ESMTP id 261752E2D78 for ; Mon, 14 Jan 2008 10:05:40 -0400 (AST) Received: from postgresql.org ([200.46.204.71]) by localhost (mx1.hub.org [200.46.204.183]) (amavisd-maia, port 10024) with ESMTP id 64914-05 for ; Mon, 14 Jan 2008 10:05:30 -0400 (AST) Received: from mx01.bfk.de (mx01.bfk.de [193.227.124.2]) by postgresql.org (Postfix) with ESMTP id 77B472E01C7 for ; Mon, 14 Jan 2008 10:05:30 -0400 (AST) Received: from mx00.int.bfk.de ([10.119.110.2]) by mx01.bfk.de with esmtps (TLS-1.0:RSA_AES_256_CBC_SHA1:32) id 1JEPwN-0003Tl-Bp for pgsql-performance@postgresql.org; Mon, 14 Jan 2008 15:05:27 +0100 Received: from fweimer by bfk.de with local id 1JEPwO-0008Sn-0T for pgsql-performance@postgresql.org; Mon, 14 Jan 2008 15:05:28 +0100 To: pgsql-performance@postgresql.org Subject: Re: Best way to index IP data? References: <13179.1200064791@sss.pgh.pa.us> <20080111162747.GT5294@mathom.us> <24044.1200082058@sss.pgh.pa.us> <20080111212703.GU5294@mathom.us> <20080111213710.GD8260@alvh.no-ip.org> <20080111220235.GX5294@mathom.us> <20080111222403.GA7527@crankycanuck.ca> <1259C7B1-C948-4326-B827-849C64B61A16@blighty.com> <827iicizg7.fsf@mid.bfk.de> <20080114111921.GE5294@mathom.us> From: Florian Weimer Date: Mon, 14 Jan 2008 15:05:27 +0100 In-Reply-To: <20080114111921.GE5294@mathom.us> (Michael Stone's message of "Mon, 14 Jan 2008 06:19:24 -0500") Message-ID: <82ir1wfqoo.fsf@mid.bfk.de> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable X-Virus-Scanned: Maia Mailguard 1.0.1 X-Archive-Number: 200801/114 X-Sequence-Number: 28353 * Michael Stone: >>Classful addressing is still part of many user interfaces, for >>instance Cisco's IOS. It's not just that the CLI accepts it, it's >>also the standard[*] output format (that is, "192.0.2.0" instead of >>"192.0.2.0/24"; if no prefix length is given, the one based on the >>class is used). > > Again, is this at all ambiguous *in the context of a host IP*? (IOW, > if I say that I recieved an IP packet from 10.0.0.1, are you really > going to ask me to clarify the netmask?) Hmm. It's an argument for a separate CIDR type, not against a host type. I agree that a host type would be helpful. I currently use check constraints to ensure no one accidentally stores data of the wrong type, which is not ideal (space is not a main concern at this point). --=20 Florian Weimer BFK edv-consulting GmbH http://www.bfk.de/ Kriegsstra=DFe 100 tel: +49-721-96201-1 D-76133 Karlsruhe fax: +49-721-96201-99