[ih] Historical Tracing from Concept to Reality over 5 decades?
jack at 3kitty.org
Tue Jul 14 12:40:50 PDT 2020
>From 1970 on for the next 7 years, I was in Lick's group at MIT, first
as a student then as research staff.
At the time, Lick was very focussed on human/computer symbiosis, i.e.,
how to have computers help humans do what they do. A major part of
that was to utilize computers in the office/business environment.
We built such a system, limited by the technology of the day, but Lick
had a vision for much more. For example, the office/business
environment included notions like approval/release procedures for
official correspondence, as well as things like notarization, escrow,
third-party verification of transmission/delivery, purchase orders,
invoices, and other such business-oriented processes. The vision was to
use the computer to automate all sorts of the traditional
business-to-business interactions, using the ARPANET as the
The vehicle for implementing this kind of system was the ARPANET, and in
particular the neonatal "mail" functionality. I recall that, at the
time, you sent an email using FTP by simply connecting to the target
system, typing "MAIL <user>" as a command to FTP, and then typing your
message, signalling the end of it by typing a line with a single period.
That was unpleasant, requiring the computer to scan every document being
sent to avoid aborting a message in process if it happened to contain a
line with just a period. Abhay Bhushan was responsible for the FTP
spec, and his office was down the hall. So after enough of my
badgering, he pushed the inclusion of the new MLFL command into FTP.
Over several years, we built a lot of that "business automation"
infrastructure. But it was limited to operation within our own system,
and not very portable since it relied on the specific PDP-10 system
(ITS) and software (PDP10 Assembler and MDL).
We made many efforts to generate enthusiasm for a "structured data"
standard for interchange across the ARPANET. But such a mechanism was
only needed for the larger office automation vision, and ARPA wasn't
funding more than a few ARPANET sites to implement such experiments.
Most people just wanted (and only needed) a simple mechanism to augment
the "mail" programs that almost every system had at the time for
communication between its users. So instead of some MTP (Message
Transmission Protocol) the much simpler SMTP (Simple Message
Transmission Protocol) was widely adopted, as an "interim" first step.
Lots of people subsequently made it useful, like Ray's adoption of @ for
While that simplicity certainly facilitated the explosion of "email"
around the world, it pretty much ended standardization efforts for more
powerful mechanisms. Fifty years later, I can see some of that kind of
functionality operating now, in various kinds of online forums, and
social media sites. But, AFAIK, all of these contemporary mechanisms
are single-site in nature. A user must connect to each individual site
to join that conversation, and "cross-posting" is often discouraged or
There is no endemic and widespread standard for supporting the more
complex interactions of Licks' vision, although technologies such as XML
and MQTT seem (to me at least) to be now filling in some of the
pieces. OTOH, market forces also seem to be splitting apart
traditional universal standards. E.g., I have lots of "mail addresses"
in my contacts list, but some of them are now only accessible through
something other than SMTP-based mail, such as Facebook, LinkedIn, et
al. I haven't found any "gateways" or magic syntax (! % etc.) to
convey messages across the boundaries. Universal email is no more.
So, Lick's vision included "electronic mail", but it was actually much
broader than the "email" we all know today. Lick's vision incorporated
"electronic mail" as one piece of the larger system of human-human
interactions using computer-computer communication over the "galactic
One likely reason that Lick's vision doesn't explicitly contain
"electronic mail" is a political one. When we were building that
system, I remember Lick (and Al Vezza) emphasizing that we were *not*
building an electronic mail system. We called it "messaging", and used
terms like "communique" rather than "mail".
The reason for this was simple. "Mail" in the US was officially and
legally the exclusive responsibility of the US Postal System. No one
was allowed to compete with the USPS. I believe Lick was acutely aware
of this issue because of his roles at both MIT and ARPA. The concern
was that someone might view an ARPANET "electronic mail" effort as an
attack on the USPS, and that might cause Congress to cut ARPA funding
for network research.
So, "complicated and cumbersome"? Yes, perhaps. But that judgement
depends on what problem you're trying to solve. Lick's vision was
significantly broader than the system we use today to carry messages
like this one. Even if he didn't ever call it "electronic mail".
On 7/6/20 3:55 PM, Dave Crocker via Internet-history wrote:
> On 7/6/2020 1:59 PM, Craig Partridge wrote:
>> On Mon, Jul 6, 2020 at 2:24 PM Dave Crocker via Internet-history
>> <internet-history at elists.isoc.org
>> <mailto:internet-history at elists.isoc.org>> wrote:
>> On the other hand, Ray's opportunistic cleverness definitely /was/
>> accidental. Especially as compared against the more elaborate
>> (cumbersome) approach others were considering. But again, that's a
>> matter of the low-level detail.
>> I think this undersells Ray's insight/genius.
> I consider "opportunistic cleverness" to be high praise.
> Besides that, my experience growing up in that environment, where
> pretty much everyone would likely be called genius by regular folk,
> was that no one used that word. I quickly noted that they simply said
> someone was clever...
>> Sometimes what makes something go from intellectual concept to
>> reality is someone figuring out how to make the something simple.
>> And Ray found a way to make email simple, and easy to use, and it
>> exploded. And then it exploded again when Vittal created MSG and
>> "Answer" [modern Reply].
> What has been interesting is seeing that some people seem to naturally
> gravitate towards powerful simplicity.
> From later discussions with Ray, it was clear to me that his process,
> back in 1971, of reacting to the surrounding discussions that were
> proposing a rather complicated, cumbersome email system -- including
> printing messages onto paper and delivering them to people's desks --
> was not an elaborate sequence of thinking through a set of issues,
> formulating careful design considerations, and engineering an
> integrated system.
> Rather it was a simple moment of the insight you cite: Just make what
> really was a tiny increment, linking two existing mechanisms of
> messaging and network file copying.
> And I don't undersell the importance of that at all.
More information about the Internet-history