From owner-freebsd-arch Thu Mar 15 17:12:58 2001 Delivered-To: freebsd-arch@freebsd.org Received: from molly.straylight.com (molly.straylight.com [209.68.199.242]) by hub.freebsd.org (Postfix) with ESMTP id 674B437B71C for ; Thu, 15 Mar 2001 17:12:55 -0800 (PST) (envelope-from jonathan@graehl.org) Received: from dickie (case.straylight.com [209.68.199.244]) by molly.straylight.com (8.11.0/8.10.0) with SMTP id f2G1CnE25596 for ; Thu, 15 Mar 2001 17:12:49 -0800 From: "Jonathan Graehl" To: "freebsd-Arch" Subject: RE: ftpd SITE MD5 and "really bad links" Date: Thu, 15 Mar 2001 17:12:29 -0800 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit X-Priority: 3 (Normal) X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2910.0) In-Reply-To: <200103152223.PAA23296@usr05.primenet.com> X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4522.1200 Importance: Normal Sender: owner-freebsd-arch@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Terry said: > > If the probability of errors (which pass 32-bit-1s-complement muster) > > on the net route between the client and FTP server is as high as once > > in a gigabyte, then SITE MD5 could save lives, not just make life > > easier for ports people. > > This is a specious argument. Such errors would occur only in the > event that the file was transferred. A SITE MD5 would come out > correct, and the error would occur anyway during subsequent file > transfer. After downloading the file (with checksum-passing errors), I compute the MD5 checksum of it. I see that it does not match the SITE MD5 value, so I log into the server again and do another "SITE MD5". If that value matches what I computed for the file, I assume there was an error in the first "SITE MD5" I got, but the file is error free. If that value differs from what I computed for the file, then I assume that I do not have the file the server wants me to have, and repeat the process. If you wanted, you could even do SITE MD5 again until it is clear you are getting the correct value. This pattern is standard; I do not understand why we need to argue it. Without SITE MD5, I would have to resort to hoping that someone had (ad-hoc) made a checksum available for me to (ad-hoc) verify. I am asking the question: what is the probability of data errors which are undetected by the TCP checksum? I have had faulty ISP equipment choke on certain byte sequences, but the TCP checksum always happened to catch whatever the error was (so the transfer would simply stall). I don't see your "4 bits in error in your lossy-compressed media file is not so bad" argument as an answer to my question, nor are all file transfers sending lossy-compressed media where you don't care about errors. -Jon To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-arch" in the body of the message