home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky talk.abortion:47754 news.admin.misc:322 comp.mail.misc:3675
- Path: sparky!uunet!gossip.pyramid.com!pyramid!weitek!nadja
- From: nadja@jetsun.weitek.COM (Nadja Adolf)
- Newsgroups: talk.abortion,news.admin.misc,comp.mail.misc
- Subject: Re: Clearly NOT a FORGERY
- Message-ID: <1992Nov12.223026.8023@jetsun.weitek.COM>
- Date: 12 Nov 92 22:30:26 GMT
- References: <1992Nov11.161634.7564@rotag.mi.org> <1992Nov12.010902.24671@jetsun.weitek.COM> <1992Nov12.070041.10433@rotag.mi.org>
- Organization: WEITEK Corporation, Sunnyvale CA
- Lines: 105
-
- In article <1992Nov12.070041.10433@rotag.mi.org> kevin@rotag.mi.org (Kevin Darcy) writes:
- >In article <1992Nov12.010902.24671@jetsun.weitek.COM> nadja@jetsun.weitek.COM (Nadja Adolf) writes:
- >>In article <1992Nov11.161634.7564@rotag.mi.org> kevin@rotag.mi.org (Kevin Darcy) writes:
- >>>In article <1992Nov11.003355.29646@jetsun.weitek.COM> nadja@jetsun.weitek.COM (Nadja Adolf) writes:
- >>
- >>Ha! If you aren't in the maps, you may not be reachable at all.
-
- >The kind of host to which I'm referring would still be reachable using its
- >fully-qualified domain name (FQDN).
-
- **********************************************
- Kevin lets us know how much he hates uucp. :-)
- **********************************************
-
- >I wouldn't call that a "smart" mailer. In fact, I'd probably get the other
- >sysadmin to fix his brain-dead software, if I ever saw that happening.
- >Additionally, if these "smart" mailers are Internet hosts, such a practice
- >may very likely qualify as a violation of RFC 1123: with regard to the "local
- >part" of a hybrid address (which is what the bangpath is considered, in
- >Internet terms) gateways are supposed to interpret that "as necessary" to
- >route the mail to its destination, and cutting off the path doesn't sound
- >like it qualifies thereunder.
-
- YO, KEVIN, WAKE UP!
- And it may well cut it off if another site has that name, or if the path
- isn't recognized. Believe it or not, there are a lot of screwed up sites
- out there, and if a node on the path is out, if you don't have a UUCP
- registry, your mail may meet the bit bucket.
-
- >By comparison, do you know of any smart mailers who cut off FQDN's? If not,
- >then the hosts in question are still reachable via their FQDN's, and don't
- >*NEED* map entries. That was my point. With a little skullduggery on the part
- >of my upstream feed, rotag.mi.org wouldn't need a map entry to get its mail
- >either. But I prefer to maintain one anyway, for added reliability, and to be
- >a good net.citizen. The point is, map entries are not CRITICAL for many folks'
- >mail routing. So it hardly makes sense to flame people, like ncsu.edu, for
- >having out-of-date map entries, or no map entries at all. What's their
- >incentive to keep a complete set of UUCP map entries up to date?
-
- There incentive is having the mail arrive.
-
- >>And the address you are describing (rotag.mi.org)
- >>is NOT a UUCP address. UUCP addresses are of the form node.uucp, or
- >>wolves.uucp. UUCP addresses are NOT the same as Internet, or ARPANET,
- >>or any other net - it's plain old-fashioned UNIX to UNIX Copy Protocol.
-
- >"Plain old-fashioned UUCP"? You mean like this:
-
- Yes. Not domain-coup addressing. :-)
-
- >Script started on Thu Nov 12 00:56:34 1992
- >/dev/ttyp0: Not owner
- >rotag% uux - -r heifetz.uucp\!rmail nadja@jetsun.weitek.com
- >bad system name: heifetz.uucp
- >uux failed ( 68 )
- >rotag% ^D
- >script done on Thu Nov 12 00:57:28 1992
-
- >Gee, I guess you were wrong about "<node>.uucp" being an address
- >comprehendable by "plain old-fashioned UUCP", huh? Surprise, surprise.
-
- Give it a rest. Try sending email to the literal address 'user@node.uucp.'
- If your tables are up to date, you'll reach my home machine, and receive
- a message telling you that you probably didn't really want to talk to
- user@node.uucp.
-
- >Trust me on this one, Nadja -- although both are valid addresses for the same
- >UUCP machine, a *lot* more mailers will be able to deliver mail reliably to
- >"<user>@rotag.mi.org", than to "<user>@rotag.uucp".
-
- No shit, Sherlock. Especially if rotag.uucp isn't on the maps. :-)
- And of course, we won't mention the importance of the records for
- reaching addresses of the FDQN form. :-)
-
- >>A message to a UUCP machine will frequently travel by Internet to the
- >>nearest UUCP gateway - and if you are on the UUCP maps, you can always
- >>send it via uunet.uu.net. And if you aren't, it may hop around, right
- >>back to the sender.
-
- >A gross oversimplification. First of all, if you're a UUCP machine, you could
- >have all the map entries in the world, and STILL not be able to get mail that
- >originates on Internet hosts, unless you have an MX record registered in the
- [more blather relating to Kevin's poor view of UUCP]
-
- Kevin justifies his failure to keep the routing maps up to date by giving
- us examples of what happens when people don't keep the maps up to date...
-
- >>There are some books on UUCP protocols. I suggest you read them.
-
- >Been there. Done that.
-
- Maybe you should take a Reading Comprehension course?
-
- > - Kevin
-
- nadja@weitek.com
-
- nadja@node.uucp
- nadja@node.com
-
- --
- The Earth Pig Bourne....
-
- nadja@node.com (prefered email address)
- nadja@weitek.com
-