Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.1 6/24/83; site hou3c.UUCP Path: utzoo!watmath!clyde!burl!hou3c!DBrown.TSDC@HI-MULTICS.ARPA From: DBrown.TSDC@HI-MULTICS.ARPA Newsgroups: net.mail.headers Subject: Re: Return path and small machines Message-ID: <275@hou3c.UUCP> Date: Tue, 14-Feb-84 18:50:00 EST Article-I.D.: hou3c.275 Posted: Tue Feb 14 18:50:00 1984 Date-Received: Fri, 17-Feb-84 04:38:21 EST Sender: ka@hou3c.UUCP (Kenneth Almquist) Lines: 22 Cc: DSullivan.CSC_SDO@HI-MULTICS, Winterton.HISCAN@HIS-PHOENIX-MULTICS, Stachour.CSCswtec@HI-MULTICS Well, it does look like something of an "envelope" issue. On one of the systems I'm familiar with there is a small host-table which takes a fully specified domain name and returns a best way to get there. Since the system in question only *has* about three ways of getting mail out, it tends to be rather trivial. The envelope gets marked "send via HI-MULTICS.ARPA" most of the time, and presumably the path it followed to get wherever it was goung reflects this, despite the fact that my real address is "Dave Brown"@CCSC-SDO.Minneapolis.Honeywell, and gets put in as "who I am". It strike me that the algorithm the oracle (which owns the host table) was published here about six months ago. By Bill Wells, maybe? In any case, it was and is a trivial name server, necessary and sufficent to get a message out on the most suitable path to a particular sub-domain. --dave DBrown.TSDC @ HI-MULTICS.ARPA watbun!drbrown @ watmath.UUCP "Dave Brown" @ CCSC-SDO.Minneapolis.Honeywell(. maybe ARPA someday)