Received: from localhost (daemon@localhost) by CS.UTK.EDU with SMTP (cf v2.9s-UTK) id SAA20142; Sat, 6 Apr 1996 18:22:49 -0500 Received: by CS.UTK.EDU (bulk_mailer v1.4); Sat, 6 Apr 1996 18:22:42 -0500 Received: from koobera.math.uic.edu (qmailr@KOOBERA.MATH.UIC.EDU [128.248.178.247]) by CS.UTK.EDU with SMTP (cf v2.9s-UTK) id SAA20120; Sat, 6 Apr 1996 18:22:40 -0500 Received: (qmail-queue invoked by uid 666); 6 Apr 1996 23:25:01 -0000 Date: 6 Apr 1996 23:25:01 -0000 Message-ID: <19960406232501.25289.qmail@koobera.math.uic.edu> From: djb@koobera.math.uic.edu (D. J. Bernstein) To: drums@cs.utk.edu Subject: Re: timezones & Date > Fine, but "sorting" is meaningless without a collating > sequence The collating sequence is Date, obviously. What we're trying to do is define Date so that, among other things, sorting by Date produces good results. It seems to me that sorting by last modification time makes more sense than sorting by mailing time. If a message has been delayed by three days, it should be at the top of my list; I don't care whether the delay was caused by the user's failure to send the message, or by his failure to dial into his ISP. (Of course, it's also possible to sort by time of receipt, but Date isn't used for that.) ---Dan