From owner-freebsd-hackers Wed Dec 11 02:26:53 1996 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.4/8.8.4) id CAA02628 for hackers-outgoing; Wed, 11 Dec 1996 02:26:53 -0800 (PST) Received: from ravenock.cybercity.dk (ravenock.cybercity.dk [194.16.57.32]) by freefall.freebsd.org (8.8.4/8.8.4) with ESMTP id CAA02615 for ; Wed, 11 Dec 1996 02:26:46 -0800 (PST) Received: (from sos@localhost) by ravenock.cybercity.dk (8.8.3/8.7.3) id LAA05791; Wed, 11 Dec 1996 11:28:43 +0100 (MET) Message-Id: <199612111028.LAA05791@ravenock.cybercity.dk> Subject: Re: PPP with network address translation In-Reply-To: <32AE4A07.167EB0E7@whistle.com> from Julian Elischer at "Dec 10, 96 09:43:35 pm" To: julian@whistle.com (Julian Elischer) Date: Wed, 11 Dec 1996 11:28:42 +0100 (MET) Cc: danny@panda.hilink.com.au, SimsS@IBM.Net, tinguely@plains.nodak.edu, freebsd-hackers@FreeBSD.ORG From: sos@FreeBSD.ORG Reply-to: sos@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk In reply to Julian Elischer who wrote: > Daniel O'Callaghan wrote: > > > > Next question with the PPP Nat, is: who is going to hack this to work > > with divert(4) sockets, so we can have a NAT box on ethernet, without the > > need for a PPP link? > > > > Danny > > that's what we do here.. > (why we wrote divert sockets actually) > but it;s inherrently kinked with the proprietary part of our system > so we can't just give it away. (we'd like to however). > but it really is a simple job. we're hoping someone just 'does it' > as our hands are tied on this one.. > > sos BTW has a purely in-kernel version of NAT (or is it HAT?) > ours is a process that works on packets that come up via the > divert sockets. HAT?? And yes I have a NAT implementation that runs entirely in kernel land, and it does all the footwork on ip, udp, tcp, ftp, irc, realaudio etc etc. It soon will support multible ip# to multible ip# and other fancy stuff. The problem is that the code is now proprietary as well, so I cannot release it either. However its not difficult to do as julian states, and I still think it should be in the kernel (or at least its needed in the apps where its used here). -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Søren Schmidt (sos@FreeBSD.org) FreeBSD Core Team Even more code to hack -- will it ever end ..