From mboxrd@z Thu Jan 1 00:00:00 1970 From: Francois Romieu Subject: Re: [Bugme-new] [Bug 28282] New: forwarding turns autoconfiguration off Date: Wed, 9 Feb 2011 08:42:21 +0100 Message-ID: <20110209074221.GA9790@electric-eye.fr.zoreil.com> References: <20110208133408.7d447e6a.akpm@linux-foundation.org> <20110208.134421.39185637.davem@davemloft.net> <4D51BFCE.2070405@msgid.danisch.de> <20110208.143046.246534138.davem@davemloft.net> <20110208224411.GA9674@electric-eye.fr.zoreil.com> <20110208234753.48abd902.billfink@mindspring.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: David Miller , hadmut@danisch.de, akpm@linux-foundation.org, netdev@vger.kernel.org, bugzilla-daemon@bugzilla.kernel.org, bugme-daemon@bugzilla.kernel.org To: Bill Fink Return-path: Received: from violet.fr.zoreil.com ([92.243.8.30]:49631 "EHLO violet.fr.zoreil.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752197Ab1BIHpd (ORCPT ); Wed, 9 Feb 2011 02:45:33 -0500 Content-Disposition: inline In-Reply-To: <20110208234753.48abd902.billfink@mindspring.com> Sender: netdev-owner@vger.kernel.org List-ID: Bill Fink : [...] > I believe there is a difference between being a router and merely > being capable of forwarding IP packets. To me, a router participates > in a routing protocol and/or advertises routes/prefixes. So perhaps > Hadmut has a valid point that autoconfiguration should not depend > on ip_forward being off, although I'm not sure what the appropriate > alternate test for not being a router should be. It is here (same document, same page): [...] 2. Terminology IP - Internet Protocol Version 6. The terms IPv4 and IPv6 are used only in contexts where necessary to avoid ambiguity. node - a device that implements IP. router - a node that forwards IP packets not explicitly addressed to itself. Why should we put our brains at pain ? Is there really a problem ? -- Ueimor