"vinton g. cerf" wrote: > ed, how would you suggest to resolve an email address if it returns ambiguous results? > Vint: 1. Short answer -- For example, by embedding a resolution rule and referring decidability to the end point, as done in this real case: -----------actual message----------------------------------------- >From aliasadm@Harris.COM Fri Apr 18 12:14:08 1997 Return-Path: aliasadm@Harris.COM Received: from sol.corp.Harris.COM (sol.corp.harris.com [137.237.104.14]) by laser.cps.softex.br (8.8.5/v3.2) with SMTP id MAA07233 for ; Fri, 18 Apr 1997 12:14:06 -0300 From: aliasadm@Harris.COM Received: by sol.corp.Harris.COM (8.6.12/Kurts Special version 2.0) id LAA14013; Fri, 18 Apr 1997 11:18:19 -0400 Date: Fri, 18 Apr 1997 11:18:19 -0400 Message-Id: <199704181518.LAA14013@sol.corp.Harris.COM> Subject: Automatic reply Status: RO X-Status: This is an automatic reply to your message addressed to: STEVEN.MORGAN@Harris.COM 2 Harris e-mail recipients are named STEVEN.MORGAN@Harris.COM Please resend your correspondence to one of the following accounts using the unique e-mail address shown on the right. MORGAN,STEVEN S CORP 1-407-724-3378 SMORGA02@harris.com MORGAN,STEVEN W ESS 1-408-734-6711 SMORGA03@harris.com ---- Original Message Follows ---- 2. Long Answer: If a name query returns two answers (and today's systems may do just that, too -- see above "short answer") but we have a meta-system that is able to disambiguate them for us or offer us a choice to do so (as we are used to, for example in multiple dictionary meanings of a word), then we are fine. The point is that rather than trying to be always decidable and complete (which WILL fail), we have a system that is as complete as possible and leaves the decidability issue to a meta-system. Different meta-systems may decide in different ways, but this is good too -- because you can choose which meta-system to use. IMO, this is just an example that the DNS needs to move from art to science. Part of this is that we need to find a logical way to describe the problems and the solutions. In other words, part of the problem in the DNS discussions, IMO, is that there is not even a language to describe the problem. Public discussions should help develop such language. And my thought is that if we are willing to use public discussions to mine the gold of truth, then we need to be open and receptive to those discussions. Cheers, Ed Gerck > > At 03:51 PM 7/30/2002 -0700, Ed Gerck wrote: > >And I am waiting to hear some substantive technical arguments why a > >name query cannot have more than one answer and still be resolvable, > >when everything else we ask in life has more than one answer and yet > >we are able to resolve them. To set the groundwork for an attempt to such > >answer, let me recall that the DNS seems to be the only logical system that > >tries to be both complete and resolvable, when we know for a mathematical > >fact that these two qualities do not go together for a system with a complexity > >larger than arithmetics. > > Vint Cerf > SVP Architecture & Technology > WorldCom > 22001 Loudoun County Parkway, F2-4115 > Ashburn, VA 20147 > 703 886 1690 (v806 1690) > 703 886 0047 fax