wb8foz@mthvax.cs.miami.edu (David Lesher) (06/22/91)
others comment about .UUCP addresses.
point 1) There are lots of declared, valid, .UUCP addresses.
Comment: If your mailer is too stupid to handle MX's and .UUCP
addresses, it should at least be smart enough to pass the job off to a
site that can do the job. That does NOT mean uunet, unless you are the
uunet customer. There does seem to be at least one Major Manufacturer
whose main gateway can't. That says lots to me about xyz.com - none
good.
point 2) There are lots of people who post .UUCP addresses that are
NOT declared in the maps.
Comment: Update your map entry. If you can't afford to post a map
entry, unplug your mail system. If at all possible, get an MX
address.
point 3) Far, far too many people post bogus, or semi-bogus addresses.
Comment: There come in several flavors.
A) Old rn that assumes your domain is .UUCP, and uses it instead of
FQDN's. Upgrade it, grep the source for .UUCP and edit it, or replace
rn with nn {blatant plug} as a newsreader.
B) If you're part of a FQD, use it. For some reason, several major
net-companies seem to post long, convoluted bang paths rather than a
simple MX FQDN, even when the MX is properly declared. I'd *love* an
explanation for this..
C) Don't count on hiding an address in your .sig. Fix your Reply-to
header.
Now can we go back to arguing about header lines ;-?
--
A host is a host from coast to coast.....wb8foz@mthvax.cs.miami.edu
& no one will talk to a host that's close............(305) 255-RTFM
Unless the host (that isn't close)......................pob 570-335
is busy, hung or dead....................................33257-0335wisner@mica.Berkeley.EDU (Bill Wisner) (06/23/91)
>Comment: If your mailer is too stupid to handle MX's and .UUCP >addresses, it should at least be smart enough to pass the job off to a >site that can do the job. This is true for MX records, since proper handling of MXes is virtually mandatory for an Internet mail host. An Internet host doesn't have to deal with .UUCP addresses, though, particularly since the "user@host.UUCP" construct has no official standing -- it's as much an underground hack as % notation. If you want to make it easy on your users, fine -- configure your mailer to punt .UUCP addresses to a gateway. But don't think it's required. Bill Wisner <wisner@mica.Berkeley.EDU> Gryphon Gang Saratoga CA 95070
nelson@sun.soe.clarkson.edu (Russ Nelson) (06/23/91)
In article <1991Jun21.214642.8746@mthvax.cs.miami.edu> wb8foz@mthvax.cs.miami.edu (David Lesher) writes:
A) Old rn that assumes your domain is .UUCP, and uses it instead of
FQDN's. Upgrade it, grep the source for .UUCP and edit it, or replace
rn with nn {blatant plug} as a newsreader.
The .UUCP that you're referring to is in rn's INIT file.
--
--russ <nelson@clutx.clarkson.edu> I'm proud to be a humble Quaker.
I am leaving the employ of Clarkson as of June 30. Hopefully this email
address will remain. If it doesn't, use nelson@gnu.ai.mit.edu.wb8foz@mthvax.cs.miami.edu (David Lesher) (06/23/91)
> An Internet host doesn't have to >deal with .UUCP addresses, though, particularly since the "user@host.UUCP" >construct has no official standing - Yep, that's true. Instead of fixing your system to handle addresses of non-royal peerage, encourage your users that have tried to mail to those addresses, and failed, to use the alternative approach. Just have them POST their mail, several times maybe, with world-wide distribution. That will work, and who could possibly complain about it, right? Suppose ATT said: We won't bother sending calls to those folks in GTE land. We've always hated GTE, and they're a pain, so to hell with them. Or the USPS said: Why should we bust our gut sending mail to Canada - those damn Canadians don't buy our stamps. [sarcasm mode, note no smily.] After all, this IS the net, where being Politically Correct is everything, and cooperation is nothing. -- A host is a host from coast to coast.....wb8foz@mthvax.cs.miami.edu & no one will talk to a host that's close............(305) 255-RTFM Unless the host (that isn't close)......................pob 570-335 is busy, hung or dead....................................33257-0335
wisner@mica.Berkeley.EDU (Bill Wisner) (06/26/91)
David Lesher: >Yep, that's true. Instead of fixing your system to handle addresses of >non-royal peerage, encourage your users that have tried to mail to >those addresses, and failed, to use the alternative approach. Just >have them POST their mail, several times maybe, with world-wide >distribution. That will work, and who could possibly complain about >it, right? >Suppose ATT said: We won't bother sending calls to those folks in GTE >land. We've always hated GTE, and they're a pain, so to hell with >them. Or the USPS said: Why should we bust our gut sending mail to >Canada - those damn Canadians don't buy our stamps. >[sarcasm mode, note no smily.] AT&T and GTE have agreed on one single standard format for telephone numbers. The USPS and Canada use the same addresses. Those seperate networks have no problem communicating. The Internet and UUCPland have agreed one one common address format, too: user@host.domain. Any UUCP host can get a valid domain fairly easily and become part of the civilized world. I'll now repeat an offer I've made before (and Karl Kleinpaste has made quite often): if any UUCP site out there wants a domain, contact me. I will personally see that it is registered, arrange for nameservers, and if necessary even provide MX forwarding. And I'll do this all for free. Bill Wisner <wisner@mica.Berkeley.EDU> Gryphon Gang Saratoga CA 95070
wb8foz@mthvax.cs.miami.edu (David Lesher) (06/26/91)
wisner@mica.Berkeley.EDU (Bill Wisner) writes: >Any UUCP host can get a valid domain fairly >easily and become part of the civilized world. >I'll now repeat an offer I've made before (and Karl Kleinpaste has >made quite often): if any UUCP site out there wants a domain, >contact me. I will personally see that it is registered, arrange >for nameservers, and if necessary even provide MX forwarding. And >I'll do this all for free. This is an outstanding offer, and one I'd like to avail myself of. I will be moving back to the DC area soon. I have an account on a UUCP site, but none on an Internet-connected one. I'm thinking up starting up my own, and need a MX host local to the DC area. Where do I call? (Anywhere in the metro DC - i.e. 703/202/301 flat rate call will do.) [For that matter, the existing .UUCP site would love one too....] -- A host is a host from coast to coast.....wb8foz@mthvax.cs.miami.edu & no one will talk to a host that's close............(305) 255-RTFM Unless the host (that isn't close)......................pob 570-335 is busy, hung or dead....................................33257-0335
fitz@wang.com (Tom Fitzgerald) (06/26/91)
wisner@mica.Berkeley.EDU (Bill Wisner) writes: > An Internet host doesn't have to > deal with .UUCP addresses, though, particularly since the "user@host.UUCP" > construct has no official standing -- it's as much an underground hack as > % notation. If you want to make it easy on your users, fine -- configure > your mailer to punt .UUCP addresses to a gateway. But don't think it's > required. Well nothing's *required* but don't understate how useful it is. Roughly 10% of the articles in the news spool here have .UUCP return addresses. A site that can't send mail to these addresses is preventing its users from responding to a whole lot of posts. --- Tom Fitzgerald Wang Labs fitz@wang.com 1-508-967-5278 Lowell MA, USA ...!uunet!wang!fitz
rhealey@kas.helios.mn.org (Rob Healey) (06/26/91)
In article <1991Jun21.214642.8746@mthvax.cs.miami.edu> wb8foz@mthvax.cs.miami.edu (David Lesher) writes: >others comment about .UUCP addresses. > >point 1) There are lots of declared, valid, .UUCP addresses. >Comment: If your mailer is too stupid to handle MX's and .UUCP >addresses, it should at least be smart enough to pass the job off to a >site that can do the job. That does NOT mean uunet, unless you are the >uunet customer. There does seem to be at least one Major Manufacturer >whose main gateway can't. That says lots to me about xyz.com - none >good. > There is NO such thing as a valid .UUCP address; PERIOD. It is NOT a valid domain name. It is used by people who don't know any better. It is also used as a hack to tell a mailer to do a pathalias lookup. Any MX site that accepts .UUCP syntax is: 1) Violating the domain name rules. 2) Violating the domain name rules. 3) Violating the domain name rules. Get the idea?? Any reasonable mailer should immediately, if not sooner, bounce a .UUCP address back from whence it came and inform the sender to either use ! addresses if they want to get to a UUCP site or else look up the REAL domain name for the site you want to reach. Hopefully, maybe in 100 years or so, people will realize that .UUCP ISN'T a valid domain name and stands a VERY good chance of being bounced or dropped at sites that can't afford to manually route 100's of misinformed UUCP sites. ex overloaded, NO, .UUCP ISN'T a valid doamin, SA, -Rob
geoff@world.std.com (Geoff Collyer) (06/27/91)
Joseph F. Young: > So, does this mean that you are going to bounce .BITNET addresses as well > as .UUCP addresses? If you do not wish your mailer to attempt to handle > such addresses, you should ship such messages an appropriate gateway host; > rejecting such mail does your users as well as the whole net a disservice > and is not going to do anything to solve the problems with the different > mail conventions among the major networks. Bitnet is a slightly different case than the UUCP network(s): Bitnet actually has a central organisation (several actually) and when Bitnet joined CREN, Bitnet promised that it was going to get rid of the ".BITNET" pseudo-domain by registering all its hosts in the Internet DNS (either with A or MX records) and getting those hosts to understand their new, Internet names (i.e. running mail systems instead of virtual card deck transports). Now that CSnet is vanishing, Bitnet officials have been heard to renege on these very sensible promises. I think the Bitnet administration should be held to the promises they made. Eventually dropping support for ".BITNET" might encourage them. -- Geoff Collyer world.std.com!geoff, uunet.uu.net!geoff
andy@jhunix.HCF.JHU.EDU (Andy S Poling) (06/28/91)
In article <b7lsgn.58z@wang.com> fitz@wang.com (Tom Fitzgerald) writes: [...] >Roughly >10% of the articles in the news spool here have .UUCP return addresses. A >site that can't send mail to these addresses is preventing its users from >responding to a whole lot of posts. Of course 10% of those .UUCP return addresses are hosts that aren't even in the blinking maps. My mailer looks up host.UUCP in our pathalias database and turns it into a full bang-path. If it's not in our pathalias database (which is updated nightly if necessary), it goes right back where it came from with the admonition that host.UUCP is not a known UUCP host. I refuse to send something as ugly as host.UUCP to anyone else. -Andy -- Andy Poling Internet: andy@jhunix.hcf.jhu.edu UNIX Systems Programmer Bitnet: ANDY@JHUNIX Homewood Academic Computing Voice: (301)338-8096 Johns Hopkins University UUCP: uunet!mimsy!aplcen!jhunix!andy