Received: from localhost (daemon@localhost) by CS.UTK.EDU with SMTP (cf v2.9s-UTK) id FAA29985; Thu, 11 Sep 1997 05:57:57 -0400 (EDT) Received: by CS.UTK.EDU (bulk_mailer v1.7); Thu, 11 Sep 1997 05:57:39 -0400 Received: by CS.UTK.EDU (cf v2.9s-UTK) id FAA29951; Thu, 11 Sep 1997 05:57:38 -0400 (EDT) Received: from bar.pilsnet.sunet.se (bar.pilsnet.sunet.se [192.36.125.14]) by CS.UTK.EDU with ESMTP (cf v2.9s-UTK) id FAA29941; Thu, 11 Sep 1997 05:57:33 -0400 (EDT) Received: (from bygg@localhost) by bar.pilsnet.sunet.se (8.8.5/8.8.5) id LAA28197 for drums@cs.utk.edu; Thu, 11 Sep 1997 11:56:18 +0200 (MET DST) Date: Thu, 11 Sep 1997 11:56:15 MET DST From: Johnny Eriksson To: drums@cs.utk.edu Subject: Re: Message-IDs and tracing In-Reply-To: Your message of 11 Sep 1997 03:02:29 -0000 Message-ID: In an unidentified msg "D. J. Bernstein" wrote: > Message-ID is an inferior substitute for the receipt; the same > Message-ID can appear on thousands of deliveries. You can mean two things here: either you mean that the same Message-ID can appear on several different messages, in which case some software is VERY broken, or you mean that when a single message gets delivered more than once, the Message-ID on the different deliveries will be the same. In the latter case, at least I think that this is exactly right. > So Message-ID logging isn't helpful. Its main effect is to waste disk > space, forcing the sysadmin to shorten the log duration. (Measure your > own compressed log files, with and without Message-IDs, if you don't > believe me.) I see nothing that prevents a disk-space-challenged sysadmin from simply stipping the Message-ID's from his log entries, especially if he don't see a use for them. > ---Dan --Johnny