jfsenior@vax1.tcd.ie (Semolina Pilchard.) (08/15/90)
People of the net.
I wonder could someone do me a favour?
I am looking for information about using uucp mail i.e. how to use 'bang -
paths' and how to use/ where to get mail maps?
Can someone point me in the way of some material I should read?
Thanks in advance...
--
J. lives at JFSENIOR@vax1.tcd.ie "I was standing next to a mountain,
chopped it down with the edge of my hand."beaulac@convex.COM (Superuser) (08/17/90)
In article <6745.26c97d33@vax1.tcd.ie> jfsenior@vax1.tcd.ie (Semolina Pilchard.) writes: >People of the net. > >I am looking for information about using uucp mail i.e. how to use 'bang - >paths' > here is a fairly simple example of using "bang paths". for mail: mail machine1\!machine2\!login for files: uucp machine1\!machine2\!/path/file the backslashes prevent the ! from being expanded. another way to accomplish this is to enclose the entire expression in single quote marks 'machine1\!machine2\!/path/file'. -bill
joe@cbnews.att.com (Joseph Judge) (08/21/90)
In article <6745.26c97d33@vax1.tcd.ie> jfsenior@vax1.tcd.ie (Semolina Pilchard.) writes: >People of the net. > >I wonder could someone do me a favour? > >I am looking for information about using uucp mail i.e. how to use 'bang - >paths' and how to use/ where to get mail maps? "bang path" stands for an addressing syntax that specifies the path from your host to a remote host and user with '!'s (bangs) used as a separator. Example: att!cblpf!joe Central.Sun.COM!jeff.jones att.com!joseph.t.judge This is different from the domain style address, which is the: user@full.host.name. Where, full.host.name is the full "dotted" domain name for the remote host and user is a local mailbox or local address. These addresses must end in a valid top-level domain. As for the maps: What you need to look at is the 'pathalias' routing database generator (by Steve Bellovin and Peter Honeyman). The info posted to comp.mail.maps is passed through pathalias, it generates a database of paths to the send hosts - relative to your site. The comp.mail.maps files are kind of boring and useless if you don't have something to compile them into hop1!hop2!hop3 What I did, in college, was have a friend at that closest "smart" site send me an ascii copy of his "paths" database. It looks like: "hosta host1!host2!host3!hosta!%s" Then, if I need to send somewhere, I grep it out, send the email to "that_smart_host!whatever_path!user" This helps overcome that big problem with using bang paths. That is, you, the user, *has* to know the full path from here to there and the connectivity is relative to your local host. Joseph Judge postmaster@ATT.COM > >Can someone point me in the way of some material I should read? > >Thanks in advance... > >-- >J. lives at JFSENIOR@vax1.tcd.ie "I was standing next to a mountain, > chopped it down with the edge of my hand."
les@chinet.chi.il.us (Leslie Mikesell) (08/22/90)
In article <1990Aug20.210800.21492@cbnews.att.com> joe@cbnews.att.com (Joseph Judge) writes: > What I did, in college, was have a friend at that closest "smart" > site send me an ascii copy of his "paths" database. > Then, if I need to send somewhere, I grep it out, send the email to > "that_smart_host!whatever_path!user" Probably a wasted effort. If the smart site is willing to forward for you it should also do the path lookup. You should have been able to send to smart_host!end_host_or_domain!user and let the mailer on the other machine do the lookup using its current paths data. Les Mikesell les@chinet.chi.il.us
joe@cbnews.att.com (Joseph Judge) (08/24/90)
In article <1990Aug21.181956.7571@chinet.chi.il.us> les@chinet.chi.il.us (Leslie Mikesell) writes: >In article <1990Aug20.210800.21492@cbnews.att.com> joe@cbnews.att.com (Joseph Judge) writes: >> What I did, in college, was have a friend at that closest "smart" >> site send me an ascii copy of his "paths" database. > >> Then, if I need to send somewhere, I grep it out, send the email to >> "that_smart_host!whatever_path!user" > >Probably a wasted effort. If the smart site is willing to forward for you >it should also do the path lookup. You should have been able to send >to smart_host!end_host_or_domain!user and let the mailer on the other >machine do the lookup using its current paths data. > >Les Mikesell > les@chinet.chi.il.us The "big project" I had at that time was routing email around the world. I wanted to specify a whole path (mail host_a!b!c!d!e!f!host_a!my_login) to wrap it around the world. At the time, they didn't have the pathalias stuff hooked into their mailer, just available to users. But, your point is good. Just find a smart site and hand it to them if you can't get pathalias and comp.mail.maps on your host. Joseph Judge postmaster@ATT.COM
emv@math.lsa.umich.edu (Edward Vielmetti) (08/26/90)
In article <1990Aug23.173359.13081@cbnews.att.com> joe@cbnews.att.com (Joseph Judge) writes:
The "big project" I had at that time was routing email around the
world.
I wanted to specify a whole path (mail host_a!b!c!d!e!f!host_a!my_login)
to wrap it around the world.
This is one reason why rabid rerouters exist and are welcomed.
--Ed
Edward Vielmetti, U of Michigan math dept <emv@math.lsa.umich.edu>
moderator, comp.archivesvjs@rhyolite.wpd.sgi.com (Vernon Schryver) (08/26/90)
In article <EMV.90Aug25175523@stag.math.lsa.umich.edu>, emv@math.lsa.umich.edu (Edward Vielmetti) writes: > > ... wanted to specify a whole path (mail host_a!b!c!d!e!f!host_a!my_login) > This is one reason why rabid rerouters exist and are welcomed. This is one reason why rabid rerouters are routinely cursed. Consider trying to debug a mail problem among b!c!d!e!f!host_a from host_a if b,c,d, or e is rabidly offisious. Do you think we could condense the next 37 rounds of this argument into 3 or 4 keywords? How about a monthly posting of the arguements of both sides? The only excuse I have for continueing is to try to limit the number of converts the Other Side makes among newcomers. I have no hope of reclaiming those who have already chosen Evil. Vernon Schryver, vjs@sgi.com
lear@turbo.bio.net (Eliot) (08/27/90)
Like I said about a month or so ago, there are some problems with Rabid Rerouting, and debugging long EMail paths is one of them. Note that it is not impossible; just that it is more difficult. In general you can look at RECEIVED lines to figure out what happened to a message. If a message is dropped without error, it gets a little more complicated, but not impossible. -- Eliot Lear [lear@turbo.bio.net]
vixie@decwrl.dec.com (Paul A Vixie) (08/27/90)
I havn't been following this discussion, so bear with me -- but if... [Eliot Lear] >> In general you can look at RECEIVED lines to figure out what happened to a >> message. If a message is dropped without error, it gets a little more >> complicated, but not impossible. Possibility of debugging approaches zero as the log files you need to peer at start to be on machines you can't log into. Also, there are a lot of mail relays that either don't put Received headers on, or they put them on in a strange order (at bottom rather than at top). There are enough reasons just in this one little area -- debugability -- to give one pause in consid- ering the question of "whether to reroute?". I still don't see that the supposed benefits of rerouting outweigh even these costs, let alone the many others I've identified here and elsewhere in years past. The main supposed advantage is that pathalias data gets out of date, and there are a lot of cheap solutions -- that is, solutions without bad side-effects like those mentioned above -- to this problem. Why reroute? (Don't worry, everybody, I'm just back from vacation. I'll go back to work shortly, this is not the beginning of another flame fest -- at least not one I plan to be part of.) -- Paul Vixie DEC Western Research Lab <vixie@wrl.dec.com> Palo Alto, California ...!decwrl!vixie