Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!wuarchive!gem.mps.ohio-state.edu!apple!claris!sts!sts!roy
From: roy@sts.sts.COM
Newsgroups: comp.mail.uucp
Subject: Re: local addresses with smail 2.5
Message-ID: <83800003@sts>
Date: 14 Nov 89 03:41:18 GMT
References: <83800001@sts>
Lines: 29
Nf-ID: #R:sts:83800001:sts:83800003:000:1351
Nf-From: sts.sts.COM!roy    Nov 13 18:51:00 1989


/* Written  2:25 pm  Nov 12, 1989 by lyndon@cs.AthabascaU.CA in sts:comp.mail.uucp */
In article <83800002@sts> roy@sts.sts.COM writes:
>Would smail3 solve 'user@sts.COM'
>vs. 'user@workstation.sts.COM' addressing anomaly?

If you want to hide several machines behind one logical name, yes
smail3 will do this. (As will sendmail, albeit nowhere near as easily).
We currently hide ~20 machines behind cs.AthabascaU.CA simply by adding
one line to the global config file.
/* End of text from sts:comp.mail.uucp */

Actually, what I want is for smail to hand anything local over to
sendmail (which, in our case, will look in /etc/hosts to deliver the
mail to the correct place over Ethernet).  As I mentioned before, it
will recognize 'user@sts.COM' as being local, but it will not
recognize 'user@workstation.sts.COM' as being anything (local or
otherwise).  Anything addressed this way to smail will be 'returned to
sender, address unknown'.  Actually, shouldn't smail recognize
'anything.sts.COM' as being local and hand it over to sendmail?  Of
course, I'm assuming there aren't any lines of the form
'wish.sts.COM = actual' ...  Since smail 2.5 apparently needs the type
of lines above, is smail3 smart enough to know that
'user@something.sts.COM' is a local address?

Roy Bixler
Asix/STS
Fremont, Cal.
roy@sts.COM -or- ...!ames!claris!sts!asix!roy