vshank%weizmann.BITNET@WISCVM.ARPA (Henry Nussbacher) (05/29/85)
For your amusement: > ----------------------- > Received: from NYU-CMCL2.ARPA by wiscvm.arpa on 05/29/85 at 08:57:34 CDT > Received: from NYU-CSD2.ARPA (nyu-csd2.arpa.ARPA) by NYU-CMCL2.ARPA; Wed, 29 > May 85 09:56:57 edt > Message-Id: <8505291356.AA29206@NYU-CMCL2.ARPA> > Received: by NYU-CSD2.ARPA; Wed, 29 May 85 09:55:11 edt > Date: Wed, 29 May 85 09:55:11 edt > From: xxxxxx@NYU-CSD2 > To: VSHANK%WEIZMANN.BITNET@WISCVM.ARPA > Subject: Re: VMNAMES & fully qualified return paths. > > > I am sorry that I have used VMNAMES from a node which does not supply a > fully qualified return address. Sorry for inconvenience. > > > However, it might be interesting for you to know that some Bitnet sites > have mailers capable of recognizing Arpanet (& other) addresses, though > they are not actually residing at gateway nodes. > A primary example of that is MAILER@CUNYVM (and, for that matter, MAILER@ > BITNIC). It would be helpful if more Bitnet nodes had this capability. > It is also very reassuring to know that, as soon as the VMNAMES program > is "migrated" to Bitnic, all the poor people from Arpanet whose mailers > don't specify a fully qualified return address will be able to get their > information painlessly. (Moreover, as Bitnic becomes an Arpanet node in the > near future, the whole procedure will be simplified considerably). > > > ------------------------- > Date: Wed, 29 May 85 17:24 IDT > From: Henry Nussbacher <vshank@weizmann> > Subject: Re: VMNAMES & fully qualified return paths. > To: <xxxxxx@NYU-CSD2.ARPA> > In-Reply-To: Your message of Wed 29 May 85 09:55:11 edt > > I accept your answer but still think NYU is wrong. What CUNYVM and > BITNIC do is a crutch and a harmful one at that. When I send a letter > from NYC to Cambridge England, I had better put in the country - England, > otherwise the postman will assume I mean Cambridge, Mass. Most postmen > will just make the letter - 'return to sender' and not try to make > "assumptions" about what country the person meant. > > This becomes very painful when a node on Arpanet has the same name as > a node on Bitnet. Example: RICE. It is a valid node in Bitnet and > a completely different machine and host in Arpanet. I don't think > Arpanet and Bitnet care to have their hostnames validated by each other. > > Still think NYU-CSD2 is better (and easier) than NYU-CSD2.ARPA? > > Hank There you have it. Please note that it is not CUNYVM's fault for accepting unqualified nodenames. It is just that since one node has the capability of doing it (and only getting it right 95% of the time - due to duplicates of nodenames and the constant problem of having to update their tables to reflect new Arpanet sites), people assume all sites in Bitnet can do it. Someone joked before about 'network police'; maybe there should be some network header police and those nodes that don't conform to standards be dropped from the Internet Host Table. Hank