demonstrated internetworking speech conferenc-
ing. Three persons located at Boston (Mas-
sachusetts), London (England) and Kjeller (Nor-
way) held a demonstration conference. The rest
of the development team, in a meeting at Uni-
versity College London at the occasion, were
solemnly listening in with the rewarding feeling
of having successively penetrated the maze of
so many difficult questions and challenges.
Each of the three sites had an LPC codec
attached to a host computer. The three comput-
ers communicated through local area nets inter-
connected through gateways, via Arpanet and
Satnet. The packet traffic in that Internet situa-
tion (new then!) was a combination of that
speech traffic together with “natural” traffic in
the Arpanet at the time. Some special knowledge
may perhaps be required to fully appreciate the
complexity of that experiment. It was one of
several major milestones during development of
Internet-technology. It took place in 1978 and
proved a number of new concepts workable.
Intricate logic functioned, and detailed prepara-
tion by several collaborating research groups
succeeded.
Standardization Procedures
One way to view the Arpanet/Internet effort is
development of new technical standards. It even
contributed to the way standardization can be
done. Traditional telecom standardization takes
place in a formal hierarchy of organizations,
mostly supported by telecom operating compa-
nies. The development of Internet-technology
led by ARPA is one prominent example of
another form of standards development.
Such international standardization is being
driven in a democratic and sometimes bureau-
cratic environment where participation and
progress are dictated as much by political moti-
vation as by technical and commercial interest.
Timeliness has sometimes suffered and stan-
dards have dropped out of pace with technical
and commercial possibilities. Some recent stan-
dardization efforts in communication and com-
puters are driven more by directly interested and
technically and commercially competent partici-
pants. There are many other examples of such
technically oriented standards development
forumstoday.
At one point in time, late 1980, some of the dif-
ferences in those two ways of standards develop-
ment were brought into focus. That was when
a meeting in the Packet Switching Protocols
Working Group (PSPWG) coincided with a
meeting of a group in The International Tele-
graph and Telephone Consultative Committee
(CCITT, now International Telecommunication
Union – Telecom Standardization Sector). That
group worked on a standard for packet switching
X.25. Impressions of the participants were
strong. We had something to learn from each
other.Although usually financed by private industry,
the success of such ad hoc forums is equally
dependent on full openness and access to the
results by anybody. Strongly competing actors
actively co-operate intimately in development- so that they can go on to compete fiercely for
their market shares and livelihood.
There have been examples of large companies
setting their own standards and keeping them to
themselves. Full openness is now considered
necessary for success of standards setting.From Arpanet to Internet
The network of nets was called Internet.One fact brought into focus and thoroughly
investigated was that various types of traffic
have different technical and economic require-
ments of the transport. Important requirements
are error freedom, time delay and economy.The development in the 1970s resulted in a num-
ber of new techniques. Especially important
were the protocols TCP and IP (Transport Con-
trol Protocol and Internet Protocol). When the
“final” standard recommendation was approved
and documented internally by the development
team, it was based on many years’ intensive
research and development.Ten groups carried out that development to-
gether as a team. It referred to itself as Packet
Switching Protocols Working Group – PSPWG.
There were eight groups in the USA, one in Eng-
land and a small group in Norway. The develop-
ment comprised investigation of a variety of
suggested methods. They were thoroughly stud-
ied theoretically and experimentally. The devel-
opment team presented and discussed intermedi-
ate results in daily communication via the new
and practical form of communication – elec-
tronic mail – and in regular meetings about
every three months. A few persons from each
group participated – typically 20 to 30 persons in
total each time.The group from ARPA, later called DARPA
(Defense Advanced Research Projects Agency),
Information Processing Techniques Office –
IPTO – led and financed the research and devel-
opment as a research project of the American
Department of Defense as basic technical
research.