home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.bbs.waffle:5695 comp.mail.uucp:2468 comp.mail.misc:4148
- Newsgroups: comp.bbs.waffle,comp.mail.uucp,comp.mail.misc
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!uwm.edu!rpi!ghost.dsi.unimi.it!univ-lyon1.fr!frmop11!psuvm!atlantis.psu.edu!barr
- From: barr@pop.psu.edu (David Barr)
- Subject: Re: Mixed format addresses
- Message-ID: <2fg1Hb16rb@atlantis.psu.edu>
- Sender: news@atlantis.psu.edu (Usenet)
- Organization: Penn State Population Research Institute
- References: <1992Dec24.185932.29142@sol.UVic.CA> <H9TkwB1w165w@willard.UUCP>
- Date: Wed, 30 Dec 92 14:58:23 GMT
- Lines: 35
-
- In article <H9TkwB1w165w@willard.UUCP> dawson@willard.UUCP (Willard Dawson) writes:
- >attempting to define mail delivery to sites that are provably on the
- >Internet. That, IMO, is rather short-sighted. "Good enough," perhaps.
- >Not, I say.
-
- What was short-sighted was providing the .UUCP hack for internet
- sites. A long-term solution would be to provide a mechanism for everyone
- to join a common mail routing system. Since global mail routing is
- inherently a multi-platform, multi-transport task, providing a common
- routing mechanism for everyone is quite difficult.
-
- The pathalias scheme has proven to be even more unreliable than
- I had ever suspected, as we will see in a few weeks when Bill Fenner
- and I post the results of a little experiment we conducted. The beauty
- in the DNS system is not in the routing aspect, which it is trivial
- since all internet sites are by nature connected to each other, but
- rather in its flexibility and distributed nature of its design.
-
- Clearly the best solution will use DNS in some fashion. That's
- why people started the .US domain, as well as provide MX'ing for other
- domains such as .COM and .ORG. IMHO this solves most of the problems.
- Granted, the MX'ing system starts to get a little inefficient the more
- hops you are away from an Internet site. Most of the sore spots can
- be handled by putting special cases in for downstream UUCP neighbors.
- (something that's probably there already). If site A is 5 hops away
- from his MX, but 3 hops away from site B, a pure MX-only sysetem will
- be worse here than a pathalias-only. (This is a hypothetical worst-case)
-
- Assume for a second that every UUCP-only site out there had a
- registered MX. What sort of other routing concerns are there?
-
- --Dave
- --
- System Administrator, Population Research Institute barr@pop.psu.edu
- #define ENOTTY 25 /* Not a typewriter */
-