[ih] What *is* "The Internet" today?

[Jack Haverty]

Interesting recent discussion about how new ideas come in to "the 
Internet".    About two years ago, I joined an ISOC working group, which 
only recently terminated after lots of sporadic discussions but few if 
any conclusions.   If anyone's curious for historical research, the 
archives of all the discussions are at 
https://elists.isoc.org/mailman/private/governance-reform/ and the 
charter is at 
https://www.internetsociety.org/board-of-trustees/governance-reform-working-group-charter/

During those two years of discussions, I learned several things, which I 
found enlightening and a bit disturbing.   But they seem relevant to the 
recent discussions here so I offer them as food for thought.   I think I 
now understand a bit more about how new ideas come in to our network 
environment, and how that's changed over the 5 or so decades of the 
Internet.

Some of you on this list might have also been on that working group.  
There were several hundred members, although most were "lurkers" and 
never said much.  I'd be curious if what I learned matches your experience.

As background, the ISOC has a Mission which is (paraphrasing here) to 
advance the Internet along 4 dimensions: global reach, openness, 
security, and trustworthiness.  Personally I've had the impression for 
quite a while now that the Internet, as I view it, has been getting 
worse in some or all of those dimensions, so I thought I'd try to help 
and joined the working group.

That led to what I learned:

1) Exactly what people mean by "the Internet" is inconsistent and the 
term itself is purposely undefined.  The ISOC views the communications 
system we all use today as being in two pieces: "the Internet", and "the 
apps" which people use.   Those apps can use the Internet to communicate.

The boundary which separates "the Internet" from "apps" is very fuzzy, 
and also purposely not defined.  So there was no consensus even in that 
working group about what the Internet is.

It seems that topics which fall into the realm of the most basic 
services, i.e., transporting IP datagrams around the planet, are the 
core elements of "the Internet".   Things that humans use, such as video 
conferencing, social media, games, telemedicine, electronic commerce, et 
al, are "apps".

The fuzzy boundary between those two is very fuzzy.  TCP seems to be 
part of "the Internet", as well as UDP, and define the core services of 
the Internet.  QUIC could be considered as another part of "the 
Internet".    Or it could be considered as an "app" simply using the 
Internet's UDP/IP functionality.   Web technology is an "app". Email 
seems like it should be an app, but the ISOC and IETF seem to have done 
a lot of work in that area.  So maybe Email is part of the Internet?  
Web technology is an app, and has it's own "society" in the W3C.

2) ISOC's Mission, and by extension the IETF's Engineering efforts, are 
focussed on making the Internet more global etc.  (I've been calling 
those the POST attributes, for Pervasive, Open, Secure, and 
Trustworthy).   Making apps POST is someone else's Mission.

3) Technical issues can "migrate" across that fuzzy boundary, and cease 
being part of the Internet.  A good example is the work done on 
packetized voice and video, which seemed (to me at least) certainly part 
of the Internet back in the early 80s.   There are many others - NNTP, 
Gopher, IRC, MBONE, etc.   Now such things are in the realm of various 
corporations who create their own proprietary technology and fight to 
have their own "silo" become the dominant one.   Such migration can be 
triggered when the activities of "the Internet" fail to produce working 
solutions (aka rough consensus and running code) in a timely fashion.   
Or even if they do, and a corporation decides to "embrace and extend" to 
create a new solution.

4) The IETF's mission (paraphrasing again) is to create high quality 
technology for "the Internet", and place it "on the shelf" for anyone 
who wishes to take and use as they see fit.   In particular, it is not 
the IETF's mission to get such technology actually deployed into the 
mechanisms we all use.   Or figure out how to perform such "heart 
transplants" on the Internet and keep it alive as such changes are 
done.  Dealing with deployment is apparently someone else's mission.

I note that this is quite unlike the early days of the Internet, when 
the job wasn't done until you could use the new technology in the live 
network.  It was quite a bit of work to make such changes happen - e.g., 
the 1983 transition which replaced NCP with the new-fangled TCP across 
the entire ARPANET.   Or replacing TCP2 with TCP4,which took months.  
Without a mission to deploy, perhaps this explains why TCPV6 still 
hasn't replaced TCPV4?

5) The IETF serves a role as a "standards body", with mechanisms to 
provide official approval for technologies that someone wants to "put on 
the shelf".   Companies can use that mechanism to get credibility for 
their in-house technology and promote their own "silo".   So you might 
find technologies "on the shelf" that were created outside the ISOC/IETF 
and brought in through the standardization process.

However, there appear to be no mechanisms for testing or certification 
of implementations of standardized technologies. That again contrasts 
with the early Internet, e.g., where there was a formal testing and 
certification procedure for TCP implementations, created and 
administered by the US National Bureau of Standards (now NIST).   Or 
even earlier the TCP testing efforts such as Jon Postel's "TCP Bakeoff" 
which achieved interoperability between different implementations of TCP.

6) Governments perceive a need to do something and regulate somehow.  
But there's a huge gap between governments and "the Internet", with all 
of those "apps" in the way.   There is no longer a strong government 
technical influence, such as was provided in the early days (80s) by 
DARPA, NSF, RSRE, NDRE, DFVLR, and other government "research" agencies.

-----------------------

 From a historical perspective, much of the ways that new ideas come 
into the Internet seems to have changed over the 5 decades of the 
Internet.  How that happened, when, and why, could make a fertile ground 
for historical analysis.

But one of the most fundamental changes has been in what "the Internet" 
means.   My personal view was derived from my experience starting with 
my indoctrination by Lick(lider) at MIT in his "galactic network" vision 
of computers communicating over some kind of "network" to help people in 
all aspects of human activity.

So, my view of "the Internet" has always been that The Internet includes 
the "apps" that people use to do whatever they do, and that the entire 
system should be POST.   But that's not apparently what "the Internet" 
is any more.  It's now just the stuff inside that fuzzy boundary.  And 
the boundary seems to be contracting as pieces migrate to become apps.

Hence my question to you -- What do *you* think "the Internet" 
includes?    How did it change over the years, and why?

Jack Haverty
MIT 1966-1977; BBN 1977-1990; Oracle 1990-1998