> "Tim Kehres" <[EMAIL PROTECTED]> wrote
>Actually, David's original assertion is indeed correct. The issue here has
>nothing to do with the reliability of TCP then vs now, but rather the state
>of email in general at the time (early 1980's). During this time, UUCP and
>the BitNET were very prominant, as well as other similar networks (CSNET,
>Fido, etc). The idea of "getting mail hop by hop to it's next nearer
>destination" was very much in the minds of most email developers at the
>time. Even in the case of destinations that were on the Internet, there
>were cases where due to network problems, a given site may not always be
>visible - hence getting it a bit closer over the net makes a lot of sense.
You are correct outside of the early Internet and ARPANET community. I
was speaking strictly about SMTP mail. UUCP which I already discussed did
do hop by hop delivery as did BITNET. Fidonet was similar. I agree email
then was more heterogeneous and there were several competing systems with
radically different architectures and delivery systems. sendmail, BTW, was
developed to deal with this rather adhoc and messy environment.
>>The RFC821 specification for SMTP simply did not contemplate SPAM
>>mail. After all, in those days it was just one happy research
>>community. Open relays were convenient, since if you wanted to deliver
>>a mail message all you needed to do was find a host running SMTP and
>>let it handle the mail routing for you.
>
>Yes and no. While it is true that the original SMTP specification did not
>contemplate SPAM (as well as several other developments), the Arpanet had
>developed to the point where it was no longer simply a research community.
>Other than the US military who were very active at the time (no surprise
>here), there were several commercial companies on the net, as well as
>several doing TCP related product development. Or at least that was what
>the group I was working with at the time thought.... :-) :-)
Yes, but all the commercial companies had to have research projects
sanctioned by ARPA to directly connect to the ARPANET. Email connectivity was a
different matter. There were UUCP, BITNET, Fidonet and CSNET gateways all over
the place. In fact with clever administration the early Internet became the
email backbone in the US. I won't go into the overseas situation, but
similar architectures were developed.
Smoot Carl-Mitchell
-
[To unsubscribe, send mail to [EMAIL PROTECTED] with
"unsubscribe firewalls" in the body of the message.]
