Message-Id: <v03100b5aaec2674454ac@resnick1.isdn.uiuc.edu>
In-Reply-To: <MailManager.849123618.370.mrc@Ikkoku-Kan.Panda.COM>
References: <v03100b44aec21ca6fe37@resnick1.isdn.uiuc.edu>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Wed, 27 Nov 1996 15:49:58 -0600
To: Mark Crispin <MRC@Panda.COM>
From: Pete Resnick <presnick@qualcomm.com>
Subject: Re: draft-ietf-drums-msg-fmt-00.txt
Cc: drums@cs.utk.edu

On 11/27/96 at 1:40 PM -0600, Mark Crispin wrote:

>The Bottom Line:
>
>It is clear from the charter that 822bis is to remain 7-bit ASCII.

OK.

>Guidance needs to be given to implementors about the pitfalls of control
>characters.  There is the Scylla of passing everything through in image form,
>and the Charybdis of damaging valid ISO-2022 based charset strings (certain
>ESC sequences, CTRL/N, and CTRL/O).

So let's talk about the document. DJB's disagreement was with SHOULD NOT
generate. Shall I leave that in or take it out in response to this?
Certainly, a note on use here would be fine, but does this impact the
syntax?

>Headers and bodies are two separate issues, and headers are more restrictive
>than bodies.  MIME has separate mechanisms for placing non-ASCII strings in
>headers and bodies.

OK.

>Implementors need to be warned about the need for NUL-safety, but it is not
>unreasonable to define NULs out of headers.

Again, what does this say about the syntax? Are you saying that NUL is a
"MAY", "SHOULD NOT" or "MUST NOT"? I originally had NUL in there but was
convinced by the conversation to take it out. I'd be glad to put it back in.

pr

--
Pete Resnick <mailto:presnick@qualcomm.com>
QUALCOMM Incorporated
Work: (217)337-6377 / Fax: (217)337-1980