From: John Stanley (stanley@peak.org)
Date: Thu Jun 06 2002 - 20:02:37 CDT
Charles Lindsey (chl@clw.cs.man.ac.uk):
> Actually, an "address" is defined as a syntactic object in RFC 2822, and
> the definition is repeated, for convenience, in the Collected Syntax
> appendix of our draft.
When it appears in an English sentence, most people are going to assume
the normal everyday meaning, not the formal one found as a puzzle at the
back of the standard.
But even so, it is interesting to compare the 2822 definition of a domain
name with the other RFCs that define legal domain names. Seems that there
are a lot of characters that are allowed in 2822 that 952 and 1035 and
1123 don't permit. And even the one RFC that does say that DNS entries can
contain any binary data (but doesn't say that about "domain names") warns
that clients may have other limits, e.g., email.
> If you don't like his policy, then don't post your news through him.
Yeah, we've heard that before. That assumes you KNOW his policy at the
moment you post, and that he's told it to you. Otherwise, you have just
had your spammable address published for the world to harvest without your
permission or your knowledge. And the site has conveniently redefined the
meaning of the Sender header and violated 8.2.2 of this standard at the
same time. What a marvelous way to write a draft.
Come on, you're almost there. You've let a MUST NOT slip into 8.2.2
referring to policy violations, so why are you so hesitant to make the
same statement at another point in the draft?