[comp.mail.sendmail] mailing from a uucp site to an internet site

djm408@tijc02.UUCP (David Marks ) (02/02/90)

How do I construct a mail path to an internet site from a uucp site
(i.e what is the syntax)? I am at a uucp site (mcnc!rti!tijc02!djm408)
and I would like to mail to someone on an internet site (kent@csuchico.edu).
We do not have software that understands domain addresses directly. I am
am told that uunet is a gateway from uucp to internet. From here the path
to uunet is: rti!mcnc!rutgers!uunet, and I am told that I can construct
a mail path of the form rti!mcnc!rutgers!uunet!SOMETHINGkent@csuchico.edu
where we don't know what to use in place of "SOMETHING." 

Please e-mail responses.

-------------------------------------------------------------------------------
 #include <standard/disclaimer.h>  |  LIFE IS NOT A MALFUNCTION! - Number 5 
    ____  ___ _  _  __ ____     ____    _ _  ___  ___   _ _   ____
    /  / /  /  | /  /  /  /      /     //// /  / /  /   / /  /___
   /  / /--/   |/  /  /  /   /  /     /  / /--/ /--\   / \      /
  ~~~~ ~  ~    ~  ~~ ~~~~    ~~~  ~  ~  ~ ~  ~ ~   ~  ~  ~  ~~~~
Pony Express:                      |    Electric Avenue:
=============                      |    ================
David J. Marks                     |    ...!mcnc!rti!tijc02!djm408
M/S 3520                           |
Texas Instruments                  |    Ma Bell:
Erwin Highway/P. O. Drawer 1255    |    ========
Johnson City, TN. 37605            |    615-461-2074

gary@dgcad.SV.DG.COM (Gary Bridgewater) (02/02/90)

In article <874@tijc02.UUCP> djm408@tijc02.UUCP (David Marks         ) writes:
>How do I construct a mail path to an internet site from a uucp site
>(i.e what is the syntax)? I am at a uucp site (mcnc!rti!tijc02!djm408)
>and I would like to mail to someone on an internet site (kent@csuchico.edu).
>We do not have software that understands domain addresses directly. I am
>am told that uunet is a gateway from uucp to internet. From here the path
>to uunet is: rti!mcnc!rutgers!uunet, and I am told that I can construct
>a mail path of the form rti!mcnc!rutgers!uunet!SOMETHINGkent@csuchico.edu
>where we don't know what to use in place of "SOMETHING." 
>Please e-mail responses.

I would e-mail, but we seem to be  awash in mail experiments here. Several of
them seem to involve an un-registered host named adam which rutgers readdresses
to the registerd host adam (a sister DG site).  These messages bounce back
from our adam here and can't get any further.

Some suggestions:

First - if you have pathalias let it create the path. Just hand it
"targethost!user" and let it do it's thing (your maps are up to date, right?).
If targethost is a unresigtered leaf of a registered host then use
"registeredhost!targethost!user".

Second - if you don't - send the mail to a smart host and let it figure the
the path from there. I don't know who, in your area , wishes to be used that
way - send mail to the postmaster on your upstream hosts and ask. Then you
can send to "xx!yy!smartmailer!targethost!user". (With a possible
"registeredhost!" before "targethost".)

Third - do not mix "!"s and "@"s in outgoing mail. It is ambiguous. You
don't know how each upstream site will interpret them, especially in long
paths.  This means that the address a!b@c can mean two things - each a
valid interpretation - 1) uucp the messages to "a" and then mail it
to user "b" on "c", and 2) mail the message to host "c" which will uucp it
to user "a" on "b".  Along the same lines - user@ahost@bhost is a real ugly
thing. The syntax, if necessary, is @bhost:user@ahost. Some use 
user%ahost@bhost.
Note that you may see addresss re-written is strange ways in the headers of
mail you get. This occurs when pairs of hosts know what the other wants and
make allowances. Don't try to do it for them in advance. Headers of
incoming messages are not good (or even, necessarily useable) examples
of mailpathing.

Last - avoid using unregistered hosts within paths unless you really know what
you are doing. Having them at the end should, of course, work. This will be
a direct fallout of my First suggestion.

I will look forward to not seeing your future attempts in my mailer-daemon
inbox :-). Flames and corrections gratefully accepted.
-- 
Gary Bridgewater, Data General Corporation, Sunnyvale California
gary@sv.dg.com or {amdahl,aeras,amdcad}!dgcad!gary
The impossible we understand right away - the obvious takes a little longer.