Received: from localhost by CS.UTK.EDU with SMTP (cf v2.9s-UTK) id CAA28290; Thu, 29 Feb 1996 02:24:19 -0500 Received: by CS.UTK.EDU (bulk_mailer v1.4); Thu, 29 Feb 1996 02:23:45 -0500 Received: from wilma.cs.utk.edu by CS.UTK.EDU with ESMTP (cf v2.9s-UTK) id CAA28228; Thu, 29 Feb 1996 02:23:26 -0500 Received: from LOCALHOST by wilma.cs.utk.edu with SMTP (cf v2.11c-UTK) id CAA16024; Thu, 29 Feb 1996 02:23:25 -0500 Message-Id: <199602290723.CAA16024@wilma.cs.utk.edu> X-URI: http://www.cs.utk.edu/~moore/ From: Keith Moore To: djb@koobera.math.uic.edu (D. J. Bernstein) cc: drums@cs.utk.edu, moore@cs.utk.edu Subject: Re: comments from a newcomer In-reply-to: Your message of "29 Feb 1996 06:40:37 GMT." <19960229064037.17741.qmail@koobera.math.uic.edu> Date: Thu, 29 Feb 1996 02:23:19 -0500 Sender: moore@cs.utk.edu > > The entire purpose of the 7-bit restriction was to prevent conflicts. > What is an SMTP relay supposed to do if it receives an 8-bit message? The correct answer is "the standards don't specify what to do with illegal input". (by SMTP relay, I assume you mean one that doesn't support 8BITMIME. Relays that support 8BITMIME are governed by the RFC that defines that extension.) Both 'be liberal with what you accept' and 'be conservative with what you send' are guidelines, not standard behavior, and a lot of people recognize now that there are some scaling problems with that simple wisdom. But back to your question. It turns out that there's no good way to solve this problem. Relaying an 8-bit message was known to cause failures or other undesirable behavior in MTAs which were perfectly compliant with the spec in only accepting 7-bit messages. Bouncing such a message -- while demonstrably legal under RFC 821 -- clearly seems undesirable. Converting it to 7-bit MIME is, as you observe, incompatible with a significant portion of the (nonstandard) installed base. None of the alternatives is good, though a case for each one can be made under certain circumstances. Requiring SMTPs to suddenly become 8-bit was felt to be unacceptably harmful to the installed base. The best hope we had was that SMTPs would migrate toward being 8-bit transparent without their being required to do so. I think that's actually happening, though these things take awhile. (Also, take care to note that the behavior of certain SMTP implementations in converting 8-bit to 7-bit MIME if and only if the message has MIME header, is not mandated by either MIME or the 8BITMIME SMTP extension... unless you care to argue that such conversions should be forbidden outright by the standard, your complaint might better be directed elsewhere.) > Which servers are those? I've checked a whole bunch of hosts, and > _none_ of them support ESMTP without saying so in the greeting line. > Why not standardize what people are already doing? Because having established a particular way of doing things with protocols that are now Full Internet Standards, we should not change those protocols unless they are demonstrably broken. -------- Take the pledge! "I do not limit my speech to satisfy the whims of Congress."