
(1)  Short note on Alexander Bain

He was the son of a crofter, and reportedly protoyped his fax using
cattle jawbones tensioned with twigs of heather.

(2)  Short on the V.20 modem standard

This was an attempt to build what was effectively a parallel
multi-frequency modem, and British Telecom attempted to use it as the
basis for their Datel 100 service with less than satisfactory
results.  The system was a failure.

(3)  Short note on V.33

This was the first modulation scheme wholly based on TCM, and is a
synchronous leased line protocol which found its way into the fax
specification.  It is virtually identical to V.17 but with only type
of training.

(4)  Extended note on standards

This extended endnote gives some flavour of the scope of the whole
topic of standardization.  I'll begin with two extremely different
interpretations, illustrated by quotations.

                         The idealistic view

The first position is that taken up by one of the leading experts in
standardisation within the ITU (the International Telecommunications
Union), who writes

"standards-makers form a small, transnational subculture held together
by technical expertise and a specialised language of their own making
... Once adopted internationally, a standard is supreme reference,
beyond challenge by lawyers and courts ... it stands on its own, a
presumably fair, equitable, and common property of mankind." (Gerd
Wallenstein, Setting Global Telecommunications Standards, Artech
House 1990, page 1).

This quotation represents what we might call the `idealistic' view of
standards.  The standards set by the ITU in the field of
telecommunications have generally been among the more successful
enterprises of their type, but the idealistic view isn't a fair
reflection of what happens in the majority of real world cases, where
although some internationally established standards are extremely
successful, others have sunk virtually without trace.  At the same
time, it is quite common for a local or proprietary solution to a
technical problem to become universally adopted and become an
effective standard without going through any kind of transnational
legitimation process at all, even where a competing and previously
verified standard already exists.

A good example of both this phenomenon at work can be seen in the
contrasting fates of two related computer languages.  ALGOL was
designed by an international American and European committee, and
while most people would agree that it was of immense significance in
the development of subsequent languages, it is not used significantly
today.  Contrast this with the fate of Pascal, which was the work of
just one person, Niklaus Wirth, who was in fact one of those involved
with the Algol-68 committee.  Despite the fact that Pascal was
originally the work of one individual and was not subject to
ratification by any standards body, it has nevertheless has become one
of the most common computer languages in international use.

The failure of Algol to make a significant impact among the mass of
working programmers has often been put down to the fact that it
wasn't ever taken up by IBM :  but on the other hand, neither was
Pascal.  Other have pointed to deficiencies in the structure of the
language itself : but all languages, including Pascal, have their
weak points, and arguing about the relative merits of languages is one
of the favourite pastimes of programmers (when they aren't changing
light bulbs).  Whatever the reasons for the contrasting fates of the
two languages might be, the fact is that the standards-makers got
this one wrong.

Adoption as a standard and specification by an international committee
does not therefore guarantee global success, while being the project
of a sole developer does not involve automatic condemnation to
remaining on the fringes of mainstream activities.  It is tempting to
explain this by arguing that the reason for the failure of some
standards is simply because they are overtaken by events, and turn
out to be technically less acceptable than some competing non-standard
; and that they can easily be replace with a successful standard by
just taking an already established way of doing things and stamping
it with official approval.

                           The Cynical view

This contrasting position is the second interpretation I'll examine.
It is taken up by an author and journalist with a background in the
data communications business, who claims that

"the word `standard' when applied to data-communications is somewhat
laughable.  Proprietary systems tend to lead the market, until
someone comes along and makes a standard out of them.  The process is
then repeated as technology moves forward faster than
standardisation, and the market is thus led by the communications
manufacturers." (Sue Schofield, Modem and Communications Guidebook,
Future Business Books 1993, page 34).

This represents the `cynical' view, which is equally far from the
truth.  The model of a standard as a series of interim technical
systems which are continually being respecified as better technology
emerges turns out to be just as simplistic.  There are innumerable
examples of innovations which fail to become standards despite their
technical superiority, and of systems which succeed even though they
are not the best of the available contenders.  The contest between
VHS and Betamax for supremacy in the VCR market is one of the more
well-known instances.  Furthermore, there are many areas where, far
from standardisation being driven by advances in technology,
manufacturers are in fact being held back by the need to conform to
existing practices.

The most famous example of this phenomenon is the QWERTY typewriter
keyboard layout, which was originally designed to slow down typing
speeds to stop keys jamming.  Despite the fact that this rationale
has long since vanished and theoretically superior keyboard layouts
have been designed, the traditional layout maintains its grip not
just on the typewriter market, but also on virtually all serious
devices with any forms of alphanumeric keyboard.  A similar situation
developed in the 1980s regarding the MSDOS/IBM standard for PCs,
which many people believe has held back the development of personal
computer for well over ten years following its introduction in 1982. 
The situation is one that is no doubt familiar to most readers.

             Standards inertia and proprietary standards

The overlong reign of the DOS PC is more about conservatism and
inertia than innovation.  Far from being manufacturer-driven, it has
been holding an entire industry in thrall.  It is quite correct to
say that technology moves forward faster than standardisation, but
the market has tended to stick with the standards rather than run
with the technology.  The fact that the existence of a standard can
hold back innovations and restrict technological innovation is
something that is recognised by many of the more experienced
standards bodies.  The key to avoiding standards becoming a dead
weight on an industry is to write extensible versions that contain
sufficient room for manufacturers for innovate and experiment while
keeping downward compatibility with a core that is compatible across
all current and future implementations.  This is always easier said
than done.

One other problem that is causing increasing problems for standards
makers is that much of the latest viable technology is not in the
public domain, and has been patented or copyrighted by organisations
which often push to have their designs included in the standard so
that they can collect royalties.  The fact that something is included
in a standard doesn't make it public property.  Some standards bodies
don't, as a matter of principle, like including any technology in
their standards that requires royalty payments and is not in the
public domain.  Since most technical advances these days are
patented, there is a clear conflict here between the needs of
standards to be state-of-the-art and the equally important imperative
that a standard be public property.

In the communications arena, this problem has come up a number of
time.  The ITU deliberately chose the compression method used for
group 3 fax because it had no active patent, and were fortunate that
they managed to get the patent on the algorithm for two-dimensional
Read coding waived before its inclusion as an extension to the T.4
fax standard.  More recent enhancements to communications standards
are generally forced to use proprietary technology.  For instance,
the V.42 bis data compression standard incorporates the
Lempel-Ziv-Welch algorithms, which are patented.  All modems
incorporating V.42 bis require royalties to be paid to IBM, Unisys
and British Telecom, who hold various portions of the patents.  Even
unofficial industry standards have been running into these problems. 
Hayes, who designed the original and many subsequent versions of the
AT set, have obtained a patent on the pause, +++, pause sequence for
returning to command mode.  This has caused problems for many
manufacturers seeking to build Hayes-compatible modems but who are
unwilling pay royalties for the use of the patented technique. 
Fortunately, none of these problems affect the fax standards.
Nevertheless, the fact that intellectual property rights can be
obtained for such a wide variety of common techniques is clearly a
complication for both those who set standards and those who implement
them, as they now have to consider the likely costs of compliance in
royalty terms.

(5)  Short note on more unofficial standards bodies (IEEE and IEC)

The review of standards bodies in this chapter has concentrated on
those working in areas related to fax. Notable bodies in related
areas which we haven't covered include the IEEE (Institution of
Electrical and Electronic Engineers) and the IEC (International
Electrotechnical Commission).

The IEEE is an American professional body actively setting standards
in many areas of computing and communications, most notably in the
field of networking. The basic standards for Ethernet and Token Ring
are IEEE 802.3 and 802.5, respectively. The IEC is an international
body dating from 1905. It works closely with both the ISO and ITU,
but is autonomous. It is the primary international authority for
standards on electrical equipment and cabling, and also on
international technical vocabulary.

(6)  Short note on FOSSIL 

The FOSSIL standard developed by the Fidonet BBS community is a good
example of an unofficial communications programming standard developed
by an unofficial body. FOSSIL is an acronym for "Fido/Opus/SEAdog
Standard Interface Layer", and was developed by an ad-hoc group of
BBS sysops (bulletin board system operators) and programmers to get
round various deficiencies in the design of access routines for
personal computer serial ports.

