[ih] IPv8...

[Jack Haverty]

On 4/18/26 14:38, Tony Li wrote:
> Hi Jack,
>
>> Somewhere in the timeline of Internet History, the notion of scenarios as drivers of technical choices must have disappeared.
> No, not at all.  In fact, it’s hard to get any solution through the IETF anymore without an independent “use cases” document.
>
> Regards,
> Tony
>
"Use Cases" and "Scenarios" are different things.   Both are needed.

My understanding of "Use Cases" is that they serve to show how some 
particular technology (protocol, algorithm, whatever) can be actually 
used in the Internet.  In other words, they are driven from the 
technology side, explaining how a technology could be used.

In contrast, "Scenarios" are driven from the end-users' perspective.  
They capture things that the end users need to be able to do, given the 
overall system of technologies that exist at the time.  The C3I scenario 
I described earlier captured one example of what the aggregate of 
military end-users needed to do, using the Internet to provide the 
communications infrastructure.   If there was a technological piece 
still missing, the scenario was not possible.

A particular technology may be useful and even necessary.  But by itself 
it is likely insufficient to actually enable any but the simplest 
end-users' scenarios.  Other technologies may also be needed before a 
particular scenario is workable.   Sometimes many technologies have to 
exist and work together.

It's also conceivable that some particular decision of a technology 
precludes ever reaching the goal of enabling a scenario.  A particular 
technology decision with a "use case" may rule out approaches to other 
technology issues that must also exist to enable the scenario.

In the C3I example I described, lots of technology advances seemed to be 
likely needed.  Routing algorithms almost certainly needed to evolve.  
Congestion and flow control likely needed changes too.  In military 
contexts, security was always a requirement.   Techniques for 
prioritizing traffic flows were likely needed.  Techniques for 
compressing large documents, voice streams, et al had to be created.  
Etc.   To meet the needs of the scenario, all the technical pieces had 
to exist and work together.

I first encountered "scenarios" while I was a student, in Professor 
Licklider's group at MIT.  Lick was my adviser and later boss for 
several years in the 1970s.  About ten years earlier, Lick had written 
memos about his vision of a "galactic network" in which computers were 
available to humans everywhere, and were somehow interconnected so that 
they could communicate with each other.

Lick's training was in psychology, so he thought from the human's 
end-user point of view.  He described the "scenario" of his "galactic 
network" vision as "computers everywhere helping humans do everything 
that humans do."  He understood that such a scenario was a bit too vague 
to serve as a specification.   But for a vision that was OK, and could 
serve to create other more detailed scenarios -- like the military ones.

In the mid-70s, one of the network research topics focussed on what came 
to be called email on the ARPANET.  I recall lots of discussions with 
Lick and many others to define relevant and more detailed scenarios.  
Lick's vision was broad, encompassing all sorts of human-human 
communication.  Emails might be short notes or massive documents.  They 
might be urgent, and require mechanisms for tracking through delivery.  
They might be multi-media, perhaps starting as text communications, 
switching to conversational voice, and even evolving into an interactive 
conferencing session using text and/or voice (video was just too hard to 
think about in the 1970s).  Interactions could be saved (such as on the 
"Datacomputer", which did exist on the ARPANET and our email system used 
it).  Long "conversations" could be related to each other as something 
like today's email "threads" and "forums".  Some communications might 
have to be private, protected from prying eyes along the way.  Some 
might need to have the author, and/or recipients, verified so that you 
could believe what you saw or heard came from where you thought it came 
from.   Some might be routed through a kind of "escrow agent", who could 
later independently testify that a document was real and had been 
authored, created, delivered, and handled at particular points in time.  
The scenario for comprehensive human-human communications was very complex.

We developed a rudimentary technical architecture for such a scenario, 
to at least serve as a starting point while computer technology advanced 
and more things became feasible.    Sadly it was probably never captured 
in any form more permanent than email archives, which are probably lost 
by now.  Lots of technologies would be needed.   It would take a while.  
  Research does.

That human-human communications architecture was shelved by ARPA in the 
mid-1970s in favor of a much simpler approach, to provide an interim 
solution that we now would all recognize as today's electronic mail.   I 
think Lick's scenario is still a good target, but I don't think anyone's 
been working on it for the last 50 years.

Anyway, I hope that explains the concept of "Scenario".....

/Jack Haverty



-------------- next part --------------
A non-text attachment was scrubbed...
Name: OpenPGP_signature.asc
Type: application/pgp-signature
Size: 665 bytes
Desc: OpenPGP digital signature
URL: <http://elists.isoc.org/pipermail/internet-history/attachments/20260418/358b80e5/attachment-0001.asc>