home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!ornl!utkcs2!emory!kd4nc!vdbsan!willard!dawson
- From: dawson@willard.UUCP
- Newsgroups: comp.mail.misc
- Subject: Re: Mixed format addresses
- Message-ID: <0HVRVB5w165w@willard.UUCP>
- Date: Mon, 14 Dec 92 07:35:44 EST
- References: <1992Dec13.170721.2862@wariat.org>
- Organization: Willard's House BBS, Atlanta, GA -- +1 (404) 664 8814
- Lines: 91
-
- Again, I'm not for a minute claiming that DNS is bad, nor am I attempting
- to tell you (that's the generic, plural you, BTW) that it doesn't have
- some very nice properties. I am trying to tell you of some problems I've
- had due to failed mail. Failure to find a route for delivery of mail,
- when routes do indeed exist, point to limitations in the software.
-
- I propose that a solution to routing mail could be arrived at that has
- a higher degree of success than is currently the case with either DNS
- or pathalias routing alone.
-
- Blindly screaming "DNS has this, this, and that, so it must be better"
- doesn't do much for me. Comparisons against alternative routing techniques
- are helpful. It's a shame that the best we can seem to come up with for
- such comparisons is pathalias.
-
- zbig@wariat.org (Zbigniew J. Tyrlik) writes:
-
- > > Easier for whom to update? You (who are on the Internet)? Those whose
- > > only connection is by UUCP? Why can we not have a system that is easy
- >
- > Either you do not understand DNS, or are simply joking. For a site
- > which is uucp only connected, and has a FQDN, it is enough to set
- > it once forever... done.
-
- Ok. How, then is this different than registration in the UUCP mapping
- project? For each site for which you agree to forward mail, you must
- re-edit your configuration file. That corresponds exactly to the update
- process for the UUCP maps, meaning that for each new connection, a forward-
- ing site should send in one update.
-
- So, perhaps that "update it once and forget it" idea doesn't count as an
- advantage of DNS over pathalias routing.
-
- > For a Internet based site, it is easier for me to update my own edition
- > of DNS, and from this monet every quesry will get updated version.
-
- Yes, it is nice that DNS gets propagated so quickly. Did I ever say that
- was bad?
-
- > With pathalias - propagation time of maps is quite long, and we have had
- > multiple occurences of maps > going to /dev/null. Claiming that updating
- > maps is easier is funny.
-
- I never claimed that updating the UUCP maps was easier, nor that it is
- more effective, than updating name servers, for Internet connected sites.
- My claim is that neither is as good as we could have.
-
- > > What tells me that DNS is not so reliable as at first it might appear are
- > > the example of incorrect routing I've seen... bounced email with "no such
- > > host" or "no such domain" when it was damn well true that the domain or
- > > host did exist, as did nameservers for them. Mostly, these examples are
- > > drawn from attempts to reply to email.
- >
- > Usually it comes to the point that site unable to respond was
- > configured totally wrong... like only local nameservers...
-
- Yes, and it happens with an amazing regularity. That only said that you
- still have to deal with operator error. No system will be immune to it,
- but you can attempt to minimize the effects of it. Any smarts built into
- either pathalias or MX forwarding, to aid admins in recovering from mis-
- configured routes? Hardly. Of course, there are scripts to test the
- validity of map files. Perhaps there are similar for
-
- > you must be kidding, man ... priviledged few on Internet...
- > Are you aware that number of hosts on Internet is a magnitute
- > bigger that hosts with uucp only connections ???
-
- Are YOU aware of how many sites connect via UUCP? Many more than actually
- get registed in either mail forwarding system, I feel sure. The advent of
- cheap, readily available software to do UUCP means that you'll see MANY
- more such sites. Just because the Internet, with its much higher costs for
- connections, is larger at this point in time, does not mean that it will
- continue to be so indefinitely.
-
- Of course, when ISDN gets tariffed for residential consumption, and all
- the people with modems rush out and buy ISDN hardware, and after some
- cheap TCP/IP over ISDN software gets out in the market, then you may see
- people moving away from their cheap UUCP connections in wholesale numbers.
-
- Yeah, right. Look, I'm holding my breath.
-
- > Sorry, Willard - your bias against Internet is quite interesting...
-
- What have I said to give you the impression I am biased against the Internet?
-
-
- --
- dawson@willard.UUCP (Willard Dawson)
- gatech!vdbsan!willard!dawson
- emory!uumind!willard!dawson
- Willard's House BBS, Atlanta, GA -- +1 (404) 664 8814
-