Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!uwm.edu!rpi!tale From: tale@cs.rpi.edu (David C Lawrence) Newsgroups: news.newusers.questions Subject: Re: UUCP Addresses Message-ID: Date: 2 Feb 90 19:27:54 GMT References: <7419@tank.uchicago.edu> Organization: Rensselaer Polytechnic Institute, Troy NY Lines: 53 In <7419@tank.uchicago.edu> jill@tank.uchicago.edu (jill holly hansen) writes: > I understand that the UUCP extension on addresses is obsolete. "Becoming obsolete" might be a better way of phrasing it. There is a strong move to geographically-specified site names but the .UUCP convention is still much too prevalent to consider it obsolete. (That is, you can consider a Sun 3/50 to be obsolete but they are still very widely used.) Geographically-specified names broken down first by country, then by state, province or other large political domain, then optionally by city or other regional name, and finally the machine name. These are specified in opposite order, such that The Well, which used to be (and still is somewhat) known as well.UUCP is now well.sf.ca.us. This name says that the site is in San Francisco, California, of the United States of America. > For example, I can't respond to a posting from someone such as > xxx@whatever.UUCP without changing it to xxx@whatever.UU.net. This is not quite right. That should only work for direct customers of UUNET. Of course that is a large number of sites, but it isn't all of them. > My question is simply, why do we continue to get UUCP extensions? Different reasons; one is that there are some (many?) UUCP sites out there that don't have registered domain names. For all I know about the maps (which isn't enough; I probably shouldn't even be posting this), all of the sites that are in the maps have geographical names which the regional map coordinators determined for them. Then again, this could just be my twisted view of the way I think it should be. > Is it the "whatever" machine's fault, or should my machine's copy of > inews or rn, etc. be able to change UUCP to something proper? It's the other machine's fault. We don't even run pathalias around Rensselaer because, to my knowledge, none of the departments do UUCP any more. It would be a fair bit of work to get us prepared to mutate other people's broken address headers for a form of transfer that we don't even use. I don't know how uchicago is set up, but it could be quite the same situation there. I just bounce my mail off of various gateways instead if I ever get a .UUCP address that bounces off my primary gateway. Unfortunately when I can't get to someone who lives down !dark!and!scary!unregistered!path!to!user I just give up after the first two bounces and moan to myself about how I wish things would work out better. Dave -- (setq mail '("tale@cs.rpi.edu" "tale@ai.mit.edu" "tale@rpitsmts.bitnet")) "Nice plant. Looks like a table cloth."