[comp.mail.uucp] Leaf Nodes

andys@ulysses.att.com (Andy Sherman) (12/12/90)

In article <RZ9PT1w163w@bluemoon.uucp> grant@bluemoon.uucp (Grant DeLorean) writes:
>andys@ulysses.att.com (Andy Sherman) writes:
>
>> Umm, I must be stupid, or something.  Yes, we are the world's biggest
>> leaf node.  But how can we be responsible for unreplyable mail?  (I
>> won't argue newspaths here).  We don't forward.  Period.  Therefore,
>> any message that is unreplyable due to att being in the path could
>> not have been sent in the first place.  Call us anti-social if you
>> like, just don't call us late to dinner.
>
> I guess I should go look at the maps before speaking,

Yes you should. :^)  Here is the entry for att


   #N	.att.com, att
   #F	att-in.att.com
   #O	AT&T
   #C	Tim Thompson, Mark Horton
   #E	postmaster@att.att.com
   #T	614-860-4100
   #P	Rm 1H-135, 6200 E. Broad, Columbus, OH 43213
   #L	39 58 54 N / 82 50 15 W
   #R	registered
   #R	also on the INTERNET
   #U	atssc bellcore bu.edu cbnews  cbnewsb cbnewsc cbnewsd cbnewse 
   #U	cbnewsg cbnewsh cbnewsi cbnewsj cbnewsk cbnewsl cbnewsm cvpath
   #U	dptg emory gladys hriso icus iuvax ll1a mcdchg motown occrsh
   #U	oucsace pacbell princeton rutgers sandoz sdl tsdiag
   #U	tut.cis.ohio-state.edu watmath westmark
   #W	postmaster@att.att.com (Tim Thompson) ; Thu May 10 10:13:20 EDT 1990
   #
   #	To contact an AT&T employee, use either
   #		full.name@att.com		Ex. john.q.doe@att.com
   #	or
   #		loginid@machine.att.com		Ex. jqd@machine.att.com
   #
   att	.att.com
   att=	att.att.com
   att=	att-in.att.com
   #
   # This line put in to reserve the name ihnp4. The name "ihnp4" is ingrained 
   # in so many people's heads (and programs) that the potential exists for 
   # e-mail to be misrouted should someone else take the name ihnp4.
   #
   att	ihnp4(DEDICATED)


> but why have
>yourself listed in the maps as connecting to other systems with a
>full map entry if you don't want mail mapped through you? Since the
>whole purpose of a map entry is to allow for shortest/best route
>mapping it seems to defeat the purpose. If you don't want to forward
>mail to other sites, either don't be fully mapped or don't list the
>systems whom you don't want to forward mail to in your map entry.
>There is no need to list everyone you talk to if you won't forward
>to them...

As you can see, this is a perfectly rational map entry for a leaf
node.  Having flamed our postmasters (internally) so many times in
the past, I'll have to complement them on this one.  As you can see,
we declare att to be the gateway into domain att.com and that's all.

I think that is the most rational entry for *anybody*.  The UUCP
maps, and the fiction of a .UUCP domain, have become a disease, or at
least a major pain in the a**.  I think domains are the only way to
control the namespace, and that's the way I run my own gateways.

>Having just locked horns with a <expletive deleted>
>from MCI over this very issue (he got mad becuase mail was being
>routed through him becuase he has/had uunet in his map entry
>as a DEMAND site) I can't keep quite just now...

I'll not touch that one......

Andy Sherman/AT&T Bell Laboratories/Murray Hill, NJ
AUDIBLE:  (201) 582-5928
READABLE: andys@ulysses.att.com  or att!ulysses!andys
What? Me speak for AT&T?  You must be joking!