[discuss] Internet: the INTER-connection of local NET-works

Mon May 5 10:14:34 UTC 2014

>At 09:00 04/05/2014, willi uebelherr wrote:
>Internet: the INTER-connection of local NET-works
>
>Dear friends,
>from the final document of NetMundial we can see, that this 
>organization has no interest to strongly support the 
>self-organization of the people for their global communication 
>systems. Rather, institutions are installed to continue the 
>principles of monopolization and representation.
>
>Now I want to submit my proposal for a real Internet in this discussion group.
>
>1) The local networks
>
>The Internet is nothing more then the connection of local, 
>independent networks. They have at least one server, which is 
>connected to the local router and this router connects to the 
>adjacent networks.
>
>These local networks have a maximum of sovereignity and 
>independenence, because they maintain all the necessary resources 
>and functions locally. These local networks are organized by the 
>local people themselves.

Dear Willi,

This is a collectively worked FSP4NET comment on your mail we 
consider as extremely positive. We will publish it on our site.

Actually, if you read the beginning of the ARPA Internetting project 
(IEN 48 
<http://www.rfc-editor.org/ien/ien48.txt>http://www.rfc-editor.org/ien/ien48.txt) 
you will see that this was the exact intent of Vint Cerf; local 
extending to global and then to multitechnological neutrality. The 
questions that arise in response to yours is the one Vint then gave 
in saying that "local" is to be understood in a "loose meaning", this 
means that the *locality* is virtual as well as, therefore, 
*globality*. The internet is by design a local Virtual Global Network 
aggregation.

Before describing what to do about it, we first need to consider why 
it is not the case and, therefore, what is to be considered/corrected.

A. Three historic problems were met:

1. The US law.

In the US, the extensions from local to global datacommunications by 
the non-ATT Bell System initiatives were illegal. The FCC found a 
solution that boils down to deciding that ATT Bell Services were 
neutral and datacommunications were non-neutral services. This 
difference, which qualified as "value-added" and then as "enhanced" 
as opposed to "basic" services, permitted licensing them in 1977. 
Deregulation further came to enlarge this position in dismantling the 
ATT Bell System. Verizon won against the FCC on neutrality because 
the FCC referred to the "basic" services based legitimate rule to 
deregulate "enhanced" services. This can only be changed by a 
revision of the 1996 Telecommunication Act by Congress. This is local 
to the US but affects the entire world if it is considered as legally 
and technically global. Not if it is considered as an aggregation of 
local "global virtualities".

2. The "BUG".

The US was a necessary business, political, and technical partner to 
the international network (which had been built around the NY/SF 
international gateways, under a non-TCP/IP US technology). This was 
used by the US industry leaders to progressively transfer the 
international network from an independent, secure, and capacity 
oriented smart fringe to fringe technology to a "Unix/IP-SuperUser" 
edged end to end architecture. As a result, the global ARPA 
Internetting project has spread all over the catenet ("the collection 
of packet networks which are connected together" cf. IEN 48), in 
retaining the VGN concept (optimizing each VGN Master's view of their 
digital world) but in applying it as an exclusive "Be Unilaterally 
Global" rule, applied by a unique single authoritative control (the 
NTIA, with the intent to contract and then transfer it to a US nexus: 
the so called "ICANN community").

3. Diversity.

The "loose meaning" of IEN 48 "local" turned out to be extremely rich 
in relation to communications technologies (i.e. "loose meaning of 
local"). IEN 48 was published in July 1978 with a twofold motivation:

3.1. to optimize "the internal technology of a data network" "for 
local operation but also permit these locally optimized nets to be 
readily interconnected into an organized catenet."
3.2. "to allow new networking technology to be introduced into the 
existing catenet while remaining functionally compatible with 
existing systems."

In June 1978, Tymshare announced the Tymnet ACT (advanced 
communication technology) that used to be for establishing every 
connection of the international network during the next eight years, 
first centered in NY/SF and then meshed in Brussels, the Hage, Paris, 
Madrid, Rome, Stockholm, etc. both with public and private networks, 
in every existing technology (including X.25/75, Swift, SITA, 
Arilines, SNA, DECNET, etc. - TCP/IP in 1984), and multilingual 
(Katakana from 1982). Without ISO presentation layer six, the TCP/IP 
technology was bound to a non-secure (cf. NSA), monolingual approach. 
Political, commercial, and lack of diversity support lead to the 
current monoprotocol set situation.

What to review?

Your approach is 100% correct. Vint Cerf recently said it too: "build 
your own internet". With your own internet comes your own governance, 
your own naming support, and your own innovation. The issue is to 
correctly identify the involved strata not to create brainware (in 
people's and machines semantic processors) confusions: this means to 
clearly identify what the internet is (the initial question of 
Nathalie Coupet), what the internet  is, and what is to be locally 
deployed to incapacitate a digital independent use, locally and globally.

1. What is the catenet? As far as we understand the author of the 
definition can confirm or object: this is the collection of packet 
networks that are connected together with a common topological plan.

2. What is the internet? This is the protocol set of what has been 
added and documented in order to transmit datagrams on an end to end 
basis through the network edges.

3. What is the namespace? This is the whole system of structured 
names of people, processes, and contents.

The problem is the confusion between them (increased by the lack of a 
presentation layer differently patched by the Web, the apps, etc.) 
that is called the "IANA functions". There is only one single IANA 
function: to provide a common repository, or protocol, for these 
separated catenet, internet, namespace parameter sources in a way 
that is most convenient to *local* independent users, when 
establishing, maintaining, and operating their *VGN*, under the 
protection of their local law and nation's precautions.

So the priority is to clarify and pragmatically accept what exists 
and to optimize the responses that it may provide to local (mobile) needs.

1. The catenet address space:

The NRO provides IP addresses that are accepted by the IXs worldwide. 
We are not going to change them globally (locally we do what we 
want). We will see that we will most probably have a limited use of 
them. For organization, protection, precaution, stability, and surety 
reasons, we would prefer them to be provided by local State regulated 
NIRs (as for car registration, civil addresses, road numbers, etc.).

2. The internet protocol set:

It was actually designed as an open minded inter-machine-netiquette 
along the IEN 48 second motivation (cf. RFC 1122). The problem is the 
edge. This is addressed by two RFCs: (1) RFC 1958 (Brian Carpenter) 
which states that everything beyond the end to end datagram 
transportation is to be located at the fringe. (2) RFC 5895 (Paul 
Hoffman and Pete Resnick) which permitted to address the diversity 
issue  [in the case of linguistic diversity (WG/IDNAbis, chaired by 
Vint Cerf)]. There were two places to locate an occurrence of 
presentation layer six (i.e. internet intelligent use):

(1) next to the end, but this broke the internet rustic resilience in 
introducing an intelligent fringe before the edge (i.e. proposing an 
intelligent internet). However, this permitted fringe control by the 
remote edge provider, i.e. a Google network intelligence control. 
Local VGN masters could not accept it.
(2) after the edge, i.e. on the user side. In that confirguration, 
the Intelligent Use Interface (IUI) operates layer six [and related 
inter-services] under the sole IUser (independent user) supervision. 
The networked "local" parts of the VGN becomes the IUser's IUIs 
internetwork (that can be scattered around the globe, trades, 
languages, and technologies). This intelligent use (IUse) 
architecture with the presentation layer on the user side (PLUS) is, 
therefore, called "Interplus".

All that local VGN Masters need is that this protocol set is 
consistent, stable, and maintained, so they can document its use in 
replacing its "MUST"s by "IS/ARE"s in their VGNIC's (virtual grlobal 
network information center) best user practices. VGN Masters are 
Internet stakeholders. External behaviors of protocols should only 
change with their global consensus.

3. The interplus stratum

This is the stratum where we can work. Since there are several IUse 
possible visions and discussing, developing, and testing them is 
transparent to the internet operations, the best approach seems to be:

3.1. To use the 
<https://www.icann.org/en/about/unique-authoritative-root>https://www.icann.org/en/about/unique-authoritative-root 
testing guidance.
3.2. Identify the different FLOSS or professional groups that are 
considering VGN tools and start a neutral mailing list.
3.3. Mutualize information and projects.

What to do?

The internet has developed on experimentation. The very basis seems 
to be to create local/personal networks and extend them. This is what 
you propose.

>2) The inter-connection of local net-works
>The Internet rests on three levels.
>
>a) connection of the adjacent local networks
>b) the regional network of regional centers
>c) the global network of regional centers
>
>The technology is based primarily on directed microwave radio links. 
>The components are manufactured locally or regionally.
>All types of data are transported. Text, graphics and speech. This 
>eliminates all separate instances for the data transport.
>The transport capacities are symmetric in principle. Thus, each 
>client can themselves act as a server.

Correct. This is where the IUI concept seems useful. It provides a 
single IP standalone gateway (can be through a VPN in order to have 
it listed in .arpa for the mail server(s)): behind the IUI this is 
another net. LISP addressing can be investigated.

>3) The IP address
>The IP address is derived from the geographical position in the 
>world coordinate system. We use 64-bit for global and 64-bit for 
>local address. Because the world coordinate system WK84 is 
>distributed asymmetrically, we should strive for a symmetrical 
>system of coordinates. Maybe it already exists.

This was proposed to the ITU but opposed by the IETF (cf. Brian 
Carpenter's mails).

>The routing (geo-routing) is based on the destination address of the 
>packet relative to the position of the router. From the distance and 
>the angle wc can easy make the decisions.
>This eliminates all institutions, which deal with the management of 
>number spaces and routing. There is no Internet governance more. It 
>is not necessary.

This may not be as simple as that. However, this is where the "road 
governance" is a good image. The roads of the world are not 
fragmented, yet there is no RCANN.

>Conclusions
>This concept rests on the responsibility to all people on our 
>planet. Only if they can operate at a most independently locally or 
>regionally level, our global communication system can arise. People 
>are important and not the institutions.

WSIS (and Sao Paulo) states that we want a "people centered" society. 
I prefer the Spanish formulation "centrada en la persona".

>It also follows, that we have to manufacture the hardware components 
>local and regional self. Any form of incapacitation of people by 
>private or public institutions is terminated. But this is only 
>possible if we determine the technology itself and organize itself. 
>We do this according to the principle: Think globally, act locally.

Limiting oneself to the PLUS (presentation (extended) layer on the 
user side), to ASAP (as a protocol) local services and applications 
on top of the Internet protocol set, and with a local use of the 
global DNS, i.e. correctly using classes and TLDs, this can be at an 
extremely low cost: software development.

Plus a local meshed network and a local shared NAS.

>At 09:33 04/05/2014, Seun Ojedeji wrote:
>Hmm... are you suggesting that internet governance should happen at 
>local level, then extend to global?
>If that is what your basic technical writeup on networks is trying 
>to communicate then I agree with that view and will like to note 
>that this is already happening.
>
>What is yet to happen is to see the proceeds from those discussions 
>turn to actions.

There are a few projects that are starting to be organized that we 
know of. Would you - or others - have one *locally*. There will be a 
presentation of VGNs in the 
<https://2014.rmll.info/>https://2014.rmll.info/ in July in 
Montpellier France by JFC/Iloth's project. If anyone wants to locally 
or remotely contribute, he/she is welcome.

FSP4NET

>many greetings in solidarity, willi uebelherr
>Quetzaltenango, Guatemala

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://1net-mail.1net.org/pipermail/discuss/attachments/20140505/caee36f4/attachment.html>