Received: from localhost (daemon@localhost) by CS.UTK.EDU with SMTP (cf v2.9s-UTK) id SAA28000; Fri, 14 Aug 1998 18:57:32 -0400 (EDT) Received: by cs.cs.utk.edu (bulk_mailer v1.11); Fri, 14 Aug 1998 18:57:23 -0400 Received: by CS.UTK.EDU (cf v2.9s-UTK) id SAA27959; Fri, 14 Aug 1998 18:57:22 -0400 (EDT) Received: from windlord.stanford.edu (windlord.Stanford.EDU [36.21.0.44]) by CS.UTK.EDU with SMTP (cf v2.9s-UTK) id SAA27941; Fri, 14 Aug 1998 18:57:17 -0400 (EDT) Received: (qmail 4958 invoked by uid 500); 14 Aug 1998 22:57:10 -0000 To: drums@cs.utk.edu Subject: Re: serious VRFY ambiguity References: From: Russ Allbery In-Reply-To: Philip Hazel's message of "Fri, 14 Aug 1998 20:16:18 +0100 (BST)" Date: 14 Aug 1998 15:57:10 -0700 Message-ID: Lines: 20 X-Mailer: Gnus v5.4.66/Emacs 19.34 List-Unsubscribe: Philip Hazel writes: > On Fri 14 Aug, Peter Sylvester wrote: >> I would say that both RCPT and VRFY can be implemented in the most >> simple way, you always say yes, and then create a non delivery notice >> later. > As far as I understand it, this thread exists because there is > apparently a question as to whether that interpretation is valid as far > as VRFY goes. Agreed. If "MUST implement VRFY" simply means that one can respond positively to all queries and generate bounces after the fact if necessary, then I have no difficulties with the standard. That is not my understanding of what is being required. -- Russ Allbery (rra@stanford.edu)