[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

still more forwarding hosts

    Date: Mon, 24 Apr 89 17:32:44 EDT
    From: MILLER@vax.cam.nbs.gov

	The Symbolics mail utilities support the syntax
	user_name%host.domain@forwarding-host to do what you've asked.  We
	use this all the time.

	 -- Chuck Fry
	   Symbolics Software Support

    Sounds plausible, but when I test it by sending a message to me on my
    sun account (fs1; which forwards eventually back to me at my symbolics)
    I get from postmaster:

    Unable to deliver letter to the following recipient:
	DNA Mail failed to verify recipient miller%fs1.cam.nbs.gov@CAMVAX.nbs.DialNet.Symbolics.COM.
    However, Is the problem with the symbolics or the vax?

You didn't say that your mail forwarding host was a non-Lispm!  You also
didn't say that your Lispm thought it was in a Dialnet-derived domain, when
other hosts at your site use a real domain.  You may wish to change your
Symbolics namespace to reflect the real domain name, since they're all
(logically, at least) on the same piece of Ethernet.  Who knows, the VAX might
even recognize its own name.

FYI, here's the path your message took to reach me:

 Received: from ELEPHANT-BUTTE.SCRC.Symbolics.COM by
  VERMITHRAX.SCH.Symbolics.COM via INTERNET with SMTP id 97109; 25 Apr 89
  06:56:58 PDT 
 Received: from STONY-BROOK.SCRC.Symbolics.COM by
  ELEPHANT-BUTTE.SCRC.Symbolics.COM via CHAOS with CHAOS-MAIL id 376963; Tue
  25-Apr-89 03:41:52 EDT
 Received: from IU.AI.SRI.COM (SRI-IU.ARPA) by STONY-BROOK.SCRC.Symbolics.COM
  via INTERNET with SMTP id 584099; 25 Apr 89 03:42:39 EDT 
 Received: from Warbucks.AI.SRI.COM by IU.AI.SRI.COM via SMTP with TCP;
	   Mon, 24 Apr 89 14:36:37-PST
 Received: from vax.cam.nbs.gov by Warbucks.AI.SRI.COM with INTERNET ;
	   Mon, 24 Apr 89 14:34:08 PDT
 Received: by vax.cam.nbs.gov id <00000227042@vax.cam.nbs.gov> ; Mon,
	   24 Apr 89 17:33:05 EDT

I.e. one hop from your VAX to SLUG@WARBUCKS.AI.SRI.COM, then redistributed
from there via Internet on the public networks, and Chaosnet within Symbolics.

    Perhaps I should be a bit more explicit about our setup:

      We've got
	1) a few Symbolics (7.2)  talking CHAOS & DNA.
	2) a VAX (CAMVAX) running VMS (4.7) and Excelan's  TCP/IP (EXOS).
	3) a VAX (ENH) running VMS and Wollengong TCP/IP. This machine is
	  `directly' connected to internet.
    So, our symbolices can only talk to themselves and CAMVAX. CAMVAX can
    talk to all local machines (but uses host tables for internet

As a Symbolics employee, of course I'd like to see you running IP-TCP (our
name for it) on your Lispms; then you could send mail "directly" to the
Internet.  Since you're not, and since your only path to the net goes through
Decnet to a VAX, it looks like you'll have to use DNA-format addresses, which
[I guess, pardon my limited experience with VAXen] might look like this:


But see below for some other possibilities.

    To send email on internet, I telnet to camvax. (if the host is not in
    the tables, I've got to forward to ENH; that probably isnt that relevent
    right now, but ultimately for some addresses I'll have to forward
    through 2 hosts (->camvax->enh) to send email from my symbolics)

    So back to the failed mail:  The symbolics obviously sent the message to
    camvax who didnt understand it. It may be that the VMS mailer took it
    instead of the EXOS daemon (?); the vms mailer certainly would not
    understand the address while the exos daemon should.  
    Or maybe somebody forgot to change the "%" into a "@"? 
    Or what?

Because you sent your test message from the Lispm to the VAX via DNA,
obviously the Decnet mailer would have handled the message before anything
else happened.  Interestingly, I found this line in this message's headers:

 X-VMS-Mail-To: EXOS%"slug@ai.sri.com"

So I suspect that even on the VAX, you have to send mail to the EXOS daemon
explicitly, and feed it addresses it can parse.  Note that the daemon's
mailbox name PRECEDES the %, completely the opposite of what I told you
earlier.  Maybe the answer is to use this format:


    So close and yet so far away...
    Thanks, Bruce.

Good luck.  I hope at least one of the suggestions above works!

 -- Chuck Fry
    Symbolics Software Support