Problems with Internet today.  


5th day of October 2009,

Don E. Sprague

This paper has three parts:
- Introduction
- The basis of the problem
- Conclusion


Introduction

It is important to note that networks aren't just lines any longer.  They now consists of  a convergence of telephone lines and
computers.  The first simple networks began in 1960.  In part, problems simply come from changes in the technology of computers
and connections.  Things changed in the past and will continue to change.  The technology used in todays global network is simply
another step in a progression. Just as IP replaced other technology,  IP will be replaced by better solution and that next solution will be
replaced.  

In 1988, when the government contracted for the delivery of an Internet, in many ways, that network was a much better information
source than today’s generic Internet that eventually came from the merger with the older larger commercial network services that
preceded the NSF’s Internet. The fair use rules kept the business activity out.  When I demonstrated the new NSF Internet and did a
search, I found good information because it wasn’t filtered with answers that were sorted based on business interests. Now when we
do searches, there are thousands more answers and they are filled with business interests as well as MORE bogus sites that might
be nefarious. Merging the government’s NSF Internet with the existing commercial global networks was the biggest change in the
history of the military Arpanet path. The merger moved the closed NSF Internet and it’s predecessor restricted government networks
from their good old days environment. However; internal government networks remain in the good old days behind the fire walls that
keep the public out. Coincidently; the commercial transition to IP actually inhibited the ongoing networking technology innovation that
comes from competition.       

The 1988 NSF outsourced Internet, anecdotally,  had it’s first business advertisement violation of the fair use rules committed by a
lawyer. Using our buffet analogy, the person brought in some food and put up a sign that said where to purchase more.  Actually, the
lawyer sent out junk E-mail.  The punishment was having thousands of people throw food at him.  Actually, thousands of people sent
nasty mail to him. The term was flaming.  What a big deal for punishment.  He advertised to hundreds of thousands of people and
received a few thousand flaming e-mails.  


The basis of problems

The basis of today’s problems with the Internet spring in part from the two paths of it’s history.  The military Arpanet path spawns a
different set of problems than those that had already been addressed in the older, larger global commercial path.  The merger of the
two paths exacerbated the problems derived from the commercialization of the closed environment that grew from the military path.  

Typically people refer the openness of the beginning of the Internet and it’s predecessor closed government networks.  Recall the
family style pitch in picnic buffet in a park.  Only family members can get into the picnic buffet network. Once a family member is allowed
inside the shelter house, anything with food on it is accessible to any family member.  If another family member has food on their plate,
any other family member can take a spoon or hand full. People can bring in any food and put it anyplace.  Who knows who brought what
isn’t on any label. In the military networks, the environment had to be even more closes in that it had to be very difficult to join. Once a
user is on one of the so called open government networks using IP or a predecessor, that user has free range access. Anecdotes of
the NSF Internet describe students at universities having unfettered access to the university routers.  The students could go to any
router to do anything. It was like the students having food fights or what ever in the buffet. They could take, move, add or change any
data. That was a very closed, yet wide open environment. The Internet we have today is still based on the design that had the wide open
pitch in buffet style problems. That historical problem not only should be addressed, it must be fixed.  

Some people refer to the older larger commercial network business as closed.  Think of it as a restaurant or store along the road or
shopping center with a food court or other restaurants and stores. Anybody can come from anyplace to go to anyplace.  Anybody can
enter the food court and sit at any open table. Some of the food suppliers are fast food while others are sit down restaurant style with a
formal menu or a buffet style.    Anybody can go to one or more of the food suppliers and purchase different things to eat.  If they want
more, they can go purchase more.  Some of the food suppliers will allow a person to refill their drink.  People can go back to get a drink
refill but they can only refill from the same supplier where they purchased the drink. That solution was always open from initial
approach to today.

From the picnic buffet and food courts analogy, we see that the military Arpanet path had a very closed environment that was wide
open.  We also see that the commercial path was wide open with restrictions.  Both paths had restrictions and both had openness.
When the government’s NSF Internet or prior networks were restricted with the fair use rules, there wasn’t any business, and, there
wasn’t an exposure to geopolitical issues. The governments networks were specifically closed to allow only specific individual user
communities.  Once a user was admitted to the network, their activity was very open.  The closed nature of the government’s NSF
Internet or prior networks hadn’t provided any need to consider the issues of openness that the business world had in it’s larger older
existing network services. Business had rules boundaries and limitations in it’s design and delivery. The commercial approach always
had a provision to allow any to any but only when authorized.  

Recall the merger activity.  We had millions of users of commercial network services.  We had a small fraction of that number using
military heritage IP. The millions convert from using richly designed languages to begin using IP. We had millions of technicians
supporting their users networking activity on languages developed by standards groups and business. IP had a few thousands of
technicians supporting much smaller user groups on a less robust language.  

All of that historical reality has been ignored.  When the two environments merged, the experience of the business world was
overshadowed by the myth of the government significance in delivering the Internet. Too much business management and technical
experience was pushed aside.  IP was lacking but the mystique of the Internet was great.  The commercial experience was lost for two
other reasons.  The newly massive wealthy from online applications had a strong voice with the military Arpanet heritage leaders in the
Internet management structures.  Third, legacy commercial business leaders looked to new people for innovation.  Legacy executives
wanted to participate in the new wealth generation and they thought their experienced people couldn’t overcome the mind set of
traditional approaches. As a result, massive changes happened in the huge 35 year old commercial networking business as it merged
or converted to IP and renamed to Internet with the paradox closed yet open military Arpanet path approach.  

Some might say the Internet is having a mid-life crisis. Actually it is more of an adolescent without parental guidance. The Internet is
more like a latch key kid.  Grown up enough to go home from school and have a key to the house but having a problem exercising good
judgement. The lack of adult guidance has exacerbate the latch key syndrome.  Consider the food court analogy again.  Before the food
court, kids would exchange paper notes with other kids.  Once they moved to the food court and shopping centers, the kids learned
about electronic devices to pass notes. The idea of passing notes wasn't new but the use of technology to pass notes was new.  

The new wave application owners are like new stores in a shopping center.  The store owner doesn't need to know about the parking
lot,  the type of stone on the side of the building, the foundation of the shopping center, the plumbing and on and on. The innovation of
the new wave store owner or new wave application is accommodated or inhibited by the accessibility and design of the shopping
center. The same applies to the global network. The application owners don't need to know about the technology between their
customers and their applications. That technology isn't just an address schema that carries some data. The new wave application
owners are benefited by the fact that they can be systems and network management novices. They don't need to know or care about
that other stuff. They should and must be free of those concerns.  They need to be able to remain focused on their business and leave
the connections to others. Today's "network architects" must have a grasp of the new wave application potential. The network architects
can't proudly boast of not having any idea of what the future will bring like the Arpanet people boasted. The architects can't boast about
waiting for guidance, again,  like the Arpanet path people boasted in the past.

Things will always change from old to new.  It is a mistake to forget or overlook what caused,  enabled,  or brought about the transition.
If we don’t know where we are going, we will get to someplace. The military Arpanet people got some place because of they could
adapt to the commercial network business environment brought to us through the guidance from the commercial path.  Now they are
working hard on technical details.  However; the hard technical detail workers still need guidance for where to take the technology.    

The need to understand basic business concerns always existed in both paths. The need for security and access control always
existed.  The lack of proper design of those controls is a big problem.  Spam and Hackers aren't good words. Junk mailers and
criminals are accurate words. That activity has always been a problem for any environment. Both family buffet and food court dining
have had to deal with criminals.  If a person purchases a drink at one food court provider, then goes to another provider and gets a refill,
that person has stolen the drink refill. The analogy applies to people who would use illegal tactics to alter an Internet access device to
enable it to get a refill from a provider they aren't authorized to use. It doesn't help to lament about the good old days to justify going to a
provider who won’t allow you to take food off their plate. That type of criminal activity must not be allowed or promoted.  The design of
the family buffet style network called IP caused the need to create fire walls that get in the way of the food court style heritage from the
older larger robust commercial design. IP had to be modified to adapt to the food court commercial approach.  As a result, simply put,
fire walls are a retro-fit to address an IP buffet style design problem.  People using the generic commercial Internet want to know that
they are the only person eating off their plate. Hence; a restriction to protect a diners plate from theft.

Many people reminisce about the good old days when the military Arpanet path began it’s closed trudge with the buffet picnic open style
dining.  The commercial problems of competition were experienced and addressed throughout the larger longer massive commercial
network services path.  There weren't new commercial issues, there are just different players and a buffet style emphasis for the
solutions to food court type problems. Every business has always competed and worked to dominate their market area. It isn't new for
the Internet, it is simply a little reality.  That has been ignored like the long history of the commercial path has been ignored.  

Business created the older larger massive commercial environment that attracted the attention of the military Arpanet path people.  The
larger massive commercial user population growth obviously must have stimulated attention in the military Arpanet path technology
workers prompting them to ignore then eliminate the fair use rules. Business always employed methods of security, performance and
garnering revenue to enable growth and innovation.  The dominance of the IP technology picnic buffet style heritage technicians is a
major inhibitor to the future of the commercial global network food court style growth. The buffet style funnel mind set stifles networking
architecture and strategy requirements innovation. The adaptive quality of the new age application owners can be used to gain an
understanding of where they may like to be in the future.  However; getting the correct answers from them takes another skill set.

One of the IP fathers said something about "No one wants a surprise bill at the end of the month." There are two things wrong with that
statement, he wasn't a father of the generic global commercial Internet we know of today. He was perhaps a grandfather or an uncle of
IP or a predecessor and the closed restricted family picnic buffet style dining without the consideration of the eventual generic
commercial  Internet shopping center and food court concept, issues, concerns and solutions. That is; he and his colleagues didn't
consider the commercial Internet activity until they decided they needed to adapt to the bigger world so their historical work could
survive.  The statement should actually say;  nobody wants to go to a restaurant and find that somebody took the food off their plate and
just left crumbs yet they must pay for full price dining.  

Some of the IP giants said they never considered the things that are being done on the Internet. That is exactly the point.  They didn't
consider the reality that was already happening in the larger global commercial "generic Internet" while they were just inventing the
Internet. The online electronic customer support and business architecture includes the words: “anything can be delivered
electronically”. The word “anything” does include videos that weren't on the minds of the team at UCLA when they began tests 40 years
ago on a predecessor of the Internet.  “Anything” from the commercial business approach includes social networking and most of the
other easy-to-use applications that have drawn more than a billion people online. It is hard to have a good futuristic technical design if
you don’t have a business plan with an inclusive vision or an idea of where you are going.



Conclusion

First: The problems with the Internet today have many of their roots in the military Arpanet path’s closed user
community limited design. The merger or conversion wasn't inclusive of enough of the design considerations from
the older massively larger commercial part of the merger. We received an adaption instead of a design.  If the NSF
1988 contract had specified a realistic architecture and strategy, they might have received a better technical solution.

Second: The skill set of the military heritage people is very good for what they do but it doesn't seem to include a
broad systems and network management and technical architecture perspective. The loss of experience from the
larger number of network players has inhibited or prevented innovation.  The network design and development
people have good skills for what they do but they are delivering solutions for something to get them someplace.  
They need to define the destination before heading down the road.  

Third: This is the biggest problem since the merger. Slowed growth simply comes from the natural maturity growth
curve. Any business approach has a growth curve.  Any technology has a growth curve.  The global network
architecture and strategy is 30 years old.  It was entering it’s last phase in the early to mid 1990s. That makes it 15
years since the last phase growth spurt that was planned or predicted to occur.  The merger or the conversion of the
commercial path and the IP path is also about 15 years old.  Think about the changes in cell phone technology in the
past 15 years. The global network business architecture is old.  The IP language is old. The commercial business
architecture and strategy completed it’s phased goals by the end of the 1990s. Since then, the military Arpanet
heritage people have adapted instead of executing a new architecture and strategy.  

I understand that there is an Internet architecture board looking at detail technology levels of IP. That is an approach
that seems to be like the good old days of government,  academic, and research military Arpanet path. The
architecture and strategy for a commercial global network should be the well defined goals and requirements of the
Internet. Development and operations teams should take the requirements from the architecture and strategy to
arrive at proper new technical executions.  It seems that we might have some development people with a pseudo title
of architect. I hope I am wrong. They seem to be living in the good old days with some of the same lack of guidance
they boast of from years ago.  Operations comes from the various ISPs.  They must participate in the design and
development activity because they must deliver the final product or service.  All of these steps and more must deliver
a product or service that attracts customers.  You see, it is a circle that always comes back to addressing the needs
of the customers.


To accelerate the arrival of the next wave of growth,  there must be a proper architecture and strategy for the next
wave.  Without that,  we will eventually get someplace and it will work some way.  I understand the that technical
details for increased address capability are defined.  So what,  we have more addresses to carry more of the same
stuff. Well,  the problems and solutions of the future need more than more addresses.  Perhaps the architecture
people are actually different from the Arpanet path people.  I hope so because my children and grandchildren could
benefit from a good architecture and strategy to accelerate a plan instead of gradually getting to some eventual
outcome.  


This approach could have been thought of by anyone in the Arpanet or pre-Internet teams.  They implemented fair use rules to
specifically exclude the any to any E-business that is characteristic of the ECS Architecture and Strategy and the Internet
Business Model we know today.  



The strategy is straight forward and very easy to understand.

– IBM was the dominate seller of computers and software.  Thousands of companies worldwide had IBM computers using SNA
for their networks.  Thousands of other companies had IBM computers or competing computers using other protocols for their
networks. Those networks didn't interconnect.

--  When a customer of IBM needed support, the System Engineer (SE) or Program Support Representative (PSR) would go to
an IBM facility, do research and load a fix on a tape.  Then the SE or PSR would drive to the customer location.

– Since the information was initially in electronic form, it could be delivered electronically if there were a connection between
IBM and the customer.

– To establish that connections between IBM and it’s customers was part A of Step 1.  

– Many of IBM’s customers were also suppliers. Connecting suppliers is part B of Step 1.

– Connecting both customers and suppliers made Electronic Customer Support a two way electronic business solution. That
was the beginning of the formal Electronic Customer Support Architecture and Strategy to deliver wide scale electronic
Business to Business communication.

– Step 2, Once IBM’s customers and suppliers were connected to a common network for electronic business with IBM, those
same customers could use the same connection to perform their own electronic business with their customers and suppliers.

– Step two was a simple expansion of the IBM to many to allow any to any.

– The cost benefit work had been completed for communication with IBM.  The subsequent Business to Business
communication was virtually cost free productivity gain.

– Step 3,  While step one and two concentrated primarily on IBM and it’s customers and suppliers, step three went after other
enterprises or smaller enterprises who weren’t or might not be IBM customers or suppliers. A person who makes belts for a
retailer might have just a PC. Their electronic business activity could also be conducted through the global network.

– Step 4 consisted of the end customer who is the general public who purchased anything from any company.  

- That leads to Step 5 which is the ultimate goal to allow the architect (and you) to communicate with those who he does
business with as well as friends and family along with access to a myriad information sources, applications or users.

16, June, 2011, Don E. Sprague

The original design of TCP/IP, browsers and other software allowed security problems.

When the developers switched from NCP to TCP/IP in 1983 for transport, they moved to a technology that was
originally designed as wide open and non-secure once you are in.
- To get on, it was supposed to be secure.
- Once you are on, it was wide open.  

That is basically the way it was until the commercial ISPs adopted TCP/IP in the mid 1990s. The developers of the
original TCP/IP lament about the good old days when the Internet was wide open.  By the time TCP/IP was adopted
by the commercial ISPs, in the mid 1990s, security enhancements had been made.  Business require the security
that has been added and is used by the commercial ISPs.

Although TCP/IP was originally designed as wide open once you were in, it has been fitted with fire walls and other
security tools. Browsers and other software is also made safer as a result of today's security software.  In addition to
the software to provide security,
each user must use caution.  

Today, I received an email from a person I know.  It was from an ID they had not been using for some time.  That was
a flag to me. It had a link to another site,  That was another flag to me. Based on the link, alone,  I sent a note to the
person on the address they use today.  The person had been attacked by a criminal. The criminal used remnants of
tools and processes originally designed into TCP/IP, browsers and other software.  

One sure thing about security,  it is a never ending process because; any time anyone designs a way to
keep things secure, someone will try to break the security.  


In a planning session with banking industry members,  they pointed out that bad checks was one of their biggest
problems that cost billions of dollars annually. They said they needed to design a process to reduce that problem.  I
suggested that they use the credit card process to check account balance and clear checks at the point of sale.  The
debit could immediately be made at the Point Of Sale. Some banks use a single card that is either a credit card or a
debit card. Today, criminals are using technology to make phony checks using real account numbers.  That makes
use of actual checks more of a problem than use of debit cards.  This just goes to show that security is a continual
process and that criminals will try to find ways to bypass new security as soon as it is implemented.  


When I worked at the IBM Information/Global Network between 1983 and 1996 one of my responsibilities was
Inter-enterprise security.  We took any security problem very serious.  We worked to enhance the inter-enterprise
Service Manager as a very secure gateway.  We also worked to make the multi-enterprise email service secure.  We
worked to design security into the network. When we switched from the IBM IN Service Manager using SNA to TCP/IP
browsers and other software we took a security step back to dirt roads. We had been using the concrete and asphalt
security of SNA and the IBM IN Service Manager with a single logon process that was solid.  

Although today's ISPs provide good security,  we still don't have a good single logon process. At the IBM Information
Network,  we recognized the problem of users having separate IDs for each of the many destination. We designed a
simple secure single logon process that was part of the Service manager.  When we switched to TCP/IP,  we lost that
security feature so I have been promoting the need for it since 1996,

I was happy to see openID until I tried it. OpenID is harder to use than the Single Logon process of the IBM IN in the
mid 1980s. After trying openID,  I could not help but to take a few minutes to write a valid requirement for a secure
single ID process.
How to have a Secure Internet

We can not have good security if we do not know who is supposed to do what.  That means, great security begins with a detailed
registration process to provide comprehensive information about each user and the connections they approve.


In the past, there were different communication methods or languages for the many Value Added networks.  They all adopted a
common communication language and became Internet Service Providers.  They did not adopt a common network management
structure.  There is a new Internet  
Secure Internet  Architecture that provides a common network management structure that all ISPs
can adopt.