Received: from localhost (daemon@localhost) by CS.UTK.EDU with SMTP (cf v2.9s-UTK) id SAA29690; Tue, 18 Aug 1998 18:39:00 -0400 (EDT) Received: by cs.cs.utk.edu (bulk_mailer v1.11); Tue, 18 Aug 1998 18:38:42 -0400 Received: by CS.UTK.EDU (cf v2.9s-UTK) id SAA29635; Tue, 18 Aug 1998 18:38:41 -0400 (EDT) Received: from koobera.math.uic.edu (koobera.math.uic.edu [131.193.178.247]) by CS.UTK.EDU with SMTP (cf v2.9s-UTK) id SAA29621; Tue, 18 Aug 1998 18:38:37 -0400 (EDT) Received: (qmail 7950 invoked by uid 666); 18 Aug 1998 22:38:52 -0000 Date: 18 Aug 1998 22:38:52 -0000 Message-ID: <19980818223852.7948.qmail@cr.yp.to> Mail-Followup-To: drums@cs.utk.edu, mailext@imc.org, ietf-submit@imc.org From: "D. J. Bernstein" To: drums@cs.utk.edu, mailext@imc.org, ietf-submit@imc.org Subject: Re: Mailrev BOF References: List-Unsubscribe: Here's one implementor's brief review of these documents. Perhaps this will save time for people not familiar with the scope of each document. Chris Newman writes: > draft-gellens-format-00.txt A method for message creators to quietly label paragraphs, including ">"-quoted paragraphs, to permit nicer displays. I'd probably watch the discussion of this if I were writing an MUA. > draft-ietf-drums-MHRegistry-03.txt A registry of header-field names. The registry is subject to IESG censorship, so it won't achieve its goal of preventing name conflicts. I have my own list at http://pobox.com/~djb/proto/immhf/index.html; please let me know if you're introducing a new field. > draft-gellens-on-demand-05.txt Sending mail to dynamic-IP dialup hosts. This is one area where ISPs are way ahead of the IETF. The Gellens document is completely out of touch with current practice; it describes an authenticated version of TURN. > draft-gellens-submit-11.txt! Message-rewriting SMTP servers. Another area where ISPs are way ahead of the IETF; see http://pobox.com/~djb/proto/ofmip.html for documentation of the real world. The Gellens document adds some religious restrictions incompatible with current practice. > draft-lindberg-anti-spam-mta-04.txt! A description of the ``anti-spam'' techniques at one site, deceptively phrased as requirements on all MTAs, mostly in violation of RFC 2119, section 6. Many portions of this document---e.g., ``fill up his disks with his own spam''---are amazingly naive; calling the whole thing ``best current practice'' is ludicrous. > draft-newman-msgheader-originfo-05.txt! A header field for recording certain types of local trace information. Standard practice is to put this information into comments in Received; there's no need for a global syntax. > draft-newman-deliver-00.txt DELIVERBY, an SMTP-link-level version of the Latest-Delivery-Time field. This is spectacularly bad engineering. > draft-freed-bsmtp-01.txt An encoding of mail messages with envelopes. I don't plan to use this; the encoding defined in http://pobox.com/~djb/proto/qmtp.txt is much easier to implement. > draft-ietf-mailext-new-fields-13.txt! Auto-Submitted, Supersedes, Expires. Auto-Submitted is like <> but without the installed base. > draft-lundblade-1pass-mult-alt-01.txt A multipart/alternative parameter to simplify some types of parsing. > draft-newman-mime-cdisp-metadata-01.txt A way to say ``Content-Disposition: No,'' for MIME parts that don't make sense by themselves. > draft-newman-auth-resp-00.txt A list of authentication errors, with a numeric code for each error. > draft-myers-sasl-pop3-05.txt New POP command, AUTH, to support SASL. No deployment plan. > draft-gellens-pop3ext-07.txt! CAPA for POP, like EHLO for SMTP. Why bother? I don't see the AUTH client stopping to ask for permission first. > draft-freed-gatesec-02.txt A discussion of one example of the obvious problems that MIME has with gateways. Are there still lots of people stuck with non-822 mailers? ---Dan Binary qmail distributions are allowed! http://pobox.com/~djb/qmail/dist.html