Re: Mandatory From field, anonymity, and hacks
Bruce Lilly <[email protected]>
| Newsgroups | gmane.ietf.rfc822 |
|---|---|
| Organization | Bruce Lilly |
| Message-ID | <[email protected]> |
On Mon January 31 2005 16:35, Charles Lindsey wrote: > > In <[email protected]> Bruce Lilly <[email protected]> writes: > > >On Sat January 29 2005 04:23, Hector Santos wrote: > >No, MTAs are not supposed to alter the message content except > >to add trace fields (Received, Return-Path). MSAs are permitted > >to make specific modifications. > > Have you ever tried sending a message with no Message-ID through sendmail? Sendmail acts as an MSA; what's your point (if you have one)? > >> Technically, a message needs no headers at all. > > >No, a message must have a message header, and that message > >header must have at least one header field. See RFC 822 > >section 4.1, RFC 2822 sections 3.5 & 3.6, and RFC 2046 > >section 5.2.1. > > I see nothing in those places that requires "at least one header field" > except for the provision that certain header fields are not optional, and > hence must be present. Duh. > But since you seem prepared to reduce the number of > required header fields, then you are in no position to object when Hector > suggests the possibility of reducing them to zero. He did not "suggest[...] the possibility"; he flatly (and incorrectly) claimed that none were needed. But, I suppose this is just another instance of the (in)famous Lindsey reading comprehension problem rearing its head. > And RFC 2046 section 5.2.1 says nothing about headers of messages. It addresses MIME-part headers. > >If that's an SMTP question, yes, because SMTP transfers > >messages, and the message format (822/2822) requires a > >header with certain fields. > > False premise, leading to false conclusion. Excuse me?