📄 ch06.htm
字号:
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<SCRIPT LANGUAGE="JavaScript">
<!--
function popUp(pPage) {
var fullURL = document.location;
var textURL = fullURL.toString();
var URLlen = textURL.length;
var lenMinusPage = textURL.lastIndexOf("/");
lenMinusPage += 1;
var fullPath = textURL.substring(0,lenMinusPage);
popUpWin = new Object() ;
figDoc= popUpWin.document;
zhtm= '<HTML><HEAD><TITLE>' + pPage + '</TITLE>';
zhtm += '</HEAD>';
zhtm += '<BODY bgcolor="#FFFFFF">';
zhtm += '<IMG SRC="' + fullPath + pPage + '">';
zhtm += '<P><B>' + pPage + '</B>';
zhtm += '</BODY></HTML>';
window.popUpWin.document.write(zhtm);
window.popUpWin.document.close();
// Johnny Jackson 4/28/98
}
//-->
</SCRIPT>
<META NAME="Author" Content="Steph Mineart">
<TITLE>Managing Multivendor Networks -- Ch 6 -- Standards</TITLE>
</HEAD>
<BODY TEXT="#000000" BGCOLOR="#FFFFFF">
<!--#exec cmd="/www/docs/ssi-bin/restricted_search.ssi"-->
<!--#exec cmd="/www/docs/ssi-bin/inc.ssi"-->
<CENTER>
<H1><IMG SRC="que.gif" tppabs="http://docs.rinet.ru:8080/MuNet/button/que.gif" WIDTH="171" HEIGHT="66" ALIGN="BOTTOM" BORDER="0"><BR>
<FONT COLOR="#000077">Managing Multivendor Networks</FONT></H1>
</CENTER>
<CENTER>
<P><A HREF="ch05.htm" tppabs="http://docs.rinet.ru:8080/MuNet/ch05/ch05.htm"><IMG SRC="previous.gif" tppabs="http://docs.rinet.ru:8080/MuNet/button/previous.gif" WIDTH="128" HEIGHT="28"
ALIGN="BOTTOM" ALT="Previous chapter" BORDER="0"></A><A HREF="ch07.htm" tppabs="http://docs.rinet.ru:8080/MuNet/ch07/ch07.htm"><IMG
SRC="next.gif" tppabs="http://docs.rinet.ru:8080/MuNet/button/next.gif" WIDTH="128" HEIGHT="28" ALIGN="BOTTOM" ALT="Next chapter"
BORDER="0"></A><A HREF="index-2.htm" tppabs="http://docs.rinet.ru:8080/MuNet/index.htm"><IMG SRC="contents.gif" tppabs="http://docs.rinet.ru:8080/MuNet/button/contents.gif" WIDTH="128"
HEIGHT="28" ALIGN="BOTTOM" ALT="Contents" BORDER="0"></A>
<HR>
</CENTER>
<CENTER>
<H1><FONT COLOR="#000077">- 6 -<BR>
Standards</FONT></H1>
</CENTER>
<UL>
<LI><A HREF="#Heading1">The Need for Standards</A>
<LI><A HREF="#Heading2">Standards Organizations</A>
<UL>
<LI><A HREF="#Heading3">American National Standards Institute (ANSI)</A>
<LI><A HREF="#Heading4">International Telecommunications Union (ITU)</A>
<LI><A HREF="#Heading5">European Computer Manufacturers Association (ECMA)</A>
<LI><A HREF="#Heading6">Electronic Industries Association (EIA)</A>
<LI><A HREF="#Heading7">Institute of Electrical and Electronics Engineers (IEEE)</A>
<LI><A HREF="#Heading8">International Standards Organization (ISO)</A>
<LI><A HREF="#Heading9">Exchange of Standards</A>
</UL>
<LI><A HREF="#Heading10">Emerging Standards</A>
<UL>
<LI><A HREF="#Heading11">Salutation Architecture</A>
<LI><A HREF="#Heading12">SONET (synchronous optical network)</A>
<LI><A HREF="#Heading13">Plug and Play Standard (Microsoft)</A>
<LI><A HREF="#Heading14">Firewire (IEEE 1394)</A>
<LI><A HREF="#Heading15">Common Desktop Environment (CDE)</A>
</UL>
</UL>
<P>
<HR SIZE="4">
</P>
<H2><A NAME="Heading1"></A><FONT COLOR="#000077">The Need for Standards</FONT></H2>
<P>ost computer manufacturers would rather not participate in standards organizations
or endorse any standards they did not develop. From the manufacturer's perspective,
little is to be gained (and much lost) from producing products conforming to external
standards; especially when these products might deviate from or conflict with the
manufacturer's own internal standards. However, if a manufacturer can set a de facto
standard, there is much to be gained. In other words, the issue is who is setting
which standards for whom.</P>
<P>The psychology behind the use of nonstandard interfaces is certainly older than
the computer industry--it is merely a different twist on the foot-in-the-door sales
approach. When used in the computer industry, however, this tactic takes on epic
proportions. First, you buy the basic system and a few user terminals. Then, you
add some more terminals and a few printers. As you expand, you add network devices
(modems, controllers, and so forth) to accommodate even more terminals and printers.
Eventually, you realize you have invested so much money in a single vendor that it
might no longer be economically feasible to consider any alternatives.</P>
<P>Fortunately, the economic impact of single-vendor sourcing has been diminished
by third-party companies. These companies provide compatible but often more economical
terminals, PCs, printers, disk drives, tape drives, and so forth. At the same time,
however, the compatibility of these third party products must remain so high that
they are not, in a technological sense, much different from the manufacturer's versions
of the same products. In other words, a third-party, 3270 -type terminal cannot
be dramatically different from an IBM 3270 terminal, or else it won't work.</P>
<P>In contrast, you can't buy a standard DEC VT220 terminal and plug it directly
into a standard IBM 3174 terminal controller. They simply weren't designed to be
interchangeable.</P>
<P>In the world of data communications, however, this story has a slightly different
twist. While it is understandable that hardware and software products from different
manufacturers cannot be interchanged like so many Lego blocks, most of the major
manufacturers eventually understood that they would have to be able to exchange information.
Because IBM took an early lead on the market, it became the focal point for the information
interchange. Any serious contender to the IBM throne has to exchange information
with IBM systems.</P>
<P>Beyond the realm of punched tape and punched cards, the first real means of data
exchange was via magnetic tape. On the negative side, this solution is inelegant,
is usually not well integrated with the mainstream applications, and requires operator
intervention. On the positive side, however, tape transport can accommodate a sizable
amount of data and does not require the permanent assignment of a systems analyst
(although this systems analyst is invariably required during the first few attempts
to define the required parameters on the tape load/unload commands).</P>
<P>Tape transport certainly has its limitations. For one thing, it is difficult to
move a tape across the country in a couple of hours. For another, because tapes are
a magnetic media, they are very susceptible to magnetic interference. Therefore,
tapes and travel mix about as well as oil and water.</P>
<P>To address the need for reliable and more timely transfer of information, data
communications-based alternatives were developed. The most popular and well-known
solution is Remote Job Entry (RJE) workstation emulation. Here, rather than relying
on tape transport, one machine emulates an IBM input or output device (each being
a part of an RJE workstation). To send data to an IBM system, the non-IBM system
emulates a remote card reader to transmit the data. To receive the data, the non-IBM
system emulates either a remote printer or remote card punch.</P>
<P>Of course, the core of this approach is the function of the RJE workstation. Without
the inception of that particular device (or set of devices), there would have been
nothing to emulate.</P>
<P>When IBM developed the RJE workstation, it was, in fact, addressing a shortcoming
in its own product line. Given the strategic and physical positioning of its mainframes
in central sites, IBM needed to accommodate large batches of information coming from
and going to remote satellite operations. To accommodate these remote sites, IBM
manufactured RJE workstations. These workstations were really a combination of devices
(a card reader, a card punch, a printer and a terminal, for example) but they were
handled as a single logical unit over a single data communications link.</P>
<P>For non-IBM manufacturers, this presented an ideal way of interfacing their systems
with IBM systems. Certainly their systems could emulate the different components
of an RJE work-station--a card reader when transmitting a file, a card punch when
receiving a file--and they could thus reliably exchange data in a real-time method.
This type of RJE emulation became so widespread that it became an industry-wide de
facto standard.</P>
<P>Furthermore, the popularity of emulating RJE became a standard that, in a sense,
transcended IBM. The most popular implementation of RJE emulation involved emulating
a 2780, a 3780, or both workstations. Because both workstations could, in fact, exchange
information with another workstation, 2780/3780 RJE emulation became an ideal mechanism
for exchanging information between any two systems--even if neither were an IBM system.
<H2><A NAME="Heading2"></A><FONT COLOR="#000077">Standards Organizations</FONT></H2>
<P>Although computer manufacturers would prefer that no standards be applied to them,
they have indeed come to realize that some standards are necessary. For example,
it is extremely convenient that most terminals use the same basic connector and that
the wiring within that connector uses specified voltages and tolerances. For one
thing, this agreement could prevent a possible explosion when a user plugs one type
of terminal into another type of connector. (For a real-life example of this, talk
to someone who has plugged an Apple laser printer set on the AppleTalk interface
into an IBM PC serial port.)</P>
<P>After computer manufacturers realized that they could not avoid standards, they
did the next best thing: Whenever possible, they started submitting their own standards
to the various standard organizations in hopes of gaining a competitive edge. For
example, IBM submitted its SDLC protocol to various organizations--it was subsequently
transformed and endorsed as HDLC (ISO standard), LAP-B (CCITT standard), and ADCCP
(ANSI standard). This is also an example of winning a battle but losing the war,
because SDLC, HDLC, LAP-B, and ADCCP are similar--but incompatible.</P>
<P>Still, this level of participation remains extremely important to computer manufacturers,
even when they do not always win a clear victory. By participating in the development
of emerging standards, they can add their two cents, look at what all their competitors
suggest, and most important, get feedback from the general scientific, and sometimes
user, communities. For the manufacturers, these benefits justify the price of admission
(even though they might secretly rather see the show close).</P>
<P>Standards organizations are often concerned about matters outside the somewhat
limited sphere of computers. In the international world, standards regulate radio,
telegraph, telephone and data communications within and between countries. In fact,
many standards organizations predate the invention of the modern computers. The International
Telecommunications Union (ITU), the parent organization of the Consultative Committee
for International Telegraph and Telephony (CCITT), for example, was founded by treaty
in 1865. Even within the more limited domain of the United States, some of the better-known
standards organizations develop standards that are far beyond the realm of data processing--for
example, ANSI also develops standards for ladders, car washes, and many other nondigital
industries.</P>
<P>Standard-producing organizations can be government-sponsored or independent. Standards
can be the by-product of computer-related associations in which the membership actively
defines and develops standards, or they can be the direct result of extensive and
intensive scientific research specifically aimed at developing a set of standards.
Then again, some organizations do not participate in the development of standards
at all but submit standards proposed by other organizations for broad approval.</P>
<P>There are thousands of organizations worldwide that participate, directly or indirectly,
in establishing standards for the computer and data communications industries. Six
of these organizations can be considered heavy hitters, capable of shaping the future
of data communications.
<H3><A NAME="Heading3"></A><FONT COLOR="#000077">American National Standards Institute
(ANSI)</FONT></H3>
<P>Although ANSI sponsors some research activities, it is primarily a clearinghouse
for other organizations that develop and submit standards. These organizations include
the Electronics Industries Association (EIA) and the Institute of Electrical and
Electronics Engineers (IEEE). ANSI was established in 1918 by a consortium of engineering
societies and government agencies, slowly evolving into its current structure. ANSI
is a nonprofit, independent organization that also serves as the U.S. representative
in the International Standards Organization (ISO). Like all of the larger standards
organizations, ANSI is divided into smaller subcommittees to focus on and study various
topics. One such committee of relative importance is the X.3 Standards Committee.</P>
<P>ANSI's X.3 Standards Committee is sponsored by the Computer and Business Equipment
Manufacturers Association (CBEMA). The scope of this committee is computer technology.
Technical Committees within the larger X.3 Standards Committee are appointed to focus
on areas of public data networks, transmission formats, and other computer related
topics.
<H3><A NAME="Heading4"></A><FONT COLOR="#000077">International Telecommunications
Union (ITU)</FONT></H3>
<P>ITU was established by treaty in 1865 to define standards in the emerging telecommunications
(that is, telegraph) industry. ITU was realigned into an agency underneath the United
Nations in 1947. Most technical topics within ITU are handled by two committees:
the Consultative Committee for International Radio (CCIR) and the Consultative Committee
for International Telegraph and Telephony (CCITT).</P>
<P>The CCITT's focus includes the areas of data communications, telematic services
(teletex, videotex, and facsimile), and Integrated Services Digital Networks (ISDN).
The CCITT is further divided into study groups that research standards within each
of those three key areas. Each study group works on its assigned topic for four years.
The CCITT has adopted and developed many standards, including the popular V.35 standard
for high-speed communications, but it is best known for the X.25 standard for public
data networks. The U.S. involvement with the CCITT is coordinated through the State
Department.
<H3><A NAME="Heading5"></A><FONT COLOR="#000077">European Computer Manufacturers
Association (ECMA)</FONT></H3>
<P>Established in 1961, ECMA voting membership is composed of European-based computer
manufacturers and has a nonvoting membership open to other parties that have marketing
or technical concerns about the European market. ECMA actively contributes to both
CCITT and ISO standards.
<H3><A NAME="Heading6"></A><FONT COLOR="#000077">Electronic Industries Association
(EIA)</FONT></H3>
<P>EIA is a trade organization founded in 1924. With respect to data communications,
EIA's prime concern is the interfacing of terminal, telecommunications and computer
equipment. Undoubtedly EIA's best known contribution to standards is the famous RS-232C
interface (although the group also engineered its replacement, the RS-449 interface).
EIA works closely with ANSI toward the development of standards; therefore, ANSI
quickly adopts many of the EIA recommendations as its own standards.
<H3><A NAME="Heading7"></A><FONT COLOR="#000077">Institute of Electrical and Electronics
Engineers (IEEE)</FONT></H3>
<P>The IEEE is an extremely large professional society in which members participate
in the development of standards that are forwarded to ANSI for approval. Like EIA,
the IEEE's relationship with ANSI is a direct path for standards to become adopted.
Specifically, the IEEE 802 series of standards (802.2, 802.3, 802.4 and 802.5) have
been adopted by many manufacturers, including HP (802.3) and IBM (802.5).
<H3><A NAME="Heading8"></A><FONT COLOR="#000077">International Standards Organization
(ISO)</FONT></H3>
<P>The ISO is a voluntary, independent organization founded in 1947 to find and define
international standards that could be agreed upon by a large number of countries.
The ISO's most significant contribution was the development of the Reference Model
for Open Systems Interconnect (OSI). The work for the OSI Reference Model began in
the late 1970s with the first drafts of the work appearing in the early 1980s. The
purpose of the model is to define a layered architecture for the development of future
standards.</P>
<P><B>The OSI Reference Model</B>  The OSI Reference Model is often seen
as a monolith much greater and far more awesome than intended by its creator, the
ISO. The OSI model is not a mandate for computer manufacturers to produce systems
that are of uniform design and use the same networking architecture. Instead, it
is a layered architecture for the design and implementation of standards that relate
to the interconnection of computer systems.</P>
<P>When the OSI model was introduced, compliance by the manufacturing community was
purely voluntary. The promises of OSI, however, were very attractive to the international
user community, so the private sector became a strong supporter of the OSI model.
Following suit, both the U.S. and United Kingdom governments have also backed the
OSI model through their respective Government OSI Profile (GOSIP) programs. With
both private and government sectors lined up behind the OSI Reference Model, the
computer manufacturers quickly stepped up to support OSI and its emerging standards.</P>
<P>Computer manufacturers generally have little direct experience or interest in
networks composed of systems from multiple vendors. It is of no great concern to
IBM engineers that IBM systems interface with those of Unisys. Nor do Digital executives
stay awake all night worrying about linking their systems to HP.</P>
<P>The user community, however, neither enjoys nor appreciates these constraints.
Their reality involves interfacing IBM systems with Digital systems, with HP systems,
with Sun systems, and so forth. The fact that most vendors offer interfaces to IBM
and UNIX-based systems as multivendor networking solutions is of little value or
solace to them.</P>
⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -