Internet Growing Pains - The Canter & Siegel story

INTERNET GROWING PAINS - By Andrew Lawrence



 When Laurence Canter decided to promote his law firm's immigration

 advisory services by posting out messages across the global

 Internet computer network, he suffered the electronic equivalent

 of being run out of town.

 

 Tens of thousands of Internet users from across the world showed

 their displeasure by swamping the Canter & Siegel company mailbox

 with angry warnings not to repeat the mailing. Their messages,

 along with a large number of genuinely interested responses to the

 advertisement, caused the computers at Canter & Siegel's

 commercial Internet providers to crash under the strain.

 

 Caught in the cross-fire, two suppliers decided not to carry

 Canter & Siegel's Internet traffic. They will probably face legal

 action. Meanwhile the law firm, unrepentant, quickly found a new

 Internet provider, and Martha Siegel, Laurence Canter's partner is

 promising more mailings and more controversy.

 

 To those who have never come into contact with the Internet,  the

 bitterness surrounding Canter & Siegel's actions is something of

 mystery. After all, the Internet, a massive web of interconnected

 computers extending across the globe, is ideally suited to

 mailings of this kind and can theoretically deliver a vast global

 audience of potential customers. When the National Science

 Foundation, the US government body which owns part of the Internet

 backbone, lifted its semi-official ban on advertising last year,

 it was almost inevitable someone would start pushing at the open

 door.

 

 But to many in the Internet community, Canter had committed a

 gross breach of 'Netiquette', the informal set of rules which

 govern behaviour on the Internet. Posting unsolicited electronic

 junk mail, or distributing advertising across the electronic

 bulletin boards (known as newsgroups) is considered by many to be

 a serious offence. "Don't even think about going down in Internet

 history this way" warns the moderator of one of the newsgroups.

 "It was utterly unconscionable behaviour", says Bill Washburn,

 executive director of the Commercial Internet Exchange

 Association, an industry association. "Obnoxious and intrusive"

 says Mike Godwin, of the Electronic Frontier Foundation.

 

 But Martha Siegel is neither deterred nor repentant. She sees the

 dispute as part of a wider struggle between the new pro-commercial

 users of the Internet and the academic community who once

 dominated the network but are now losing control. "If you cut

 through all this what you will find is a group of old timers who

 don't want their private domain invaded".

 

 That domain is certainly being invaded. The Internet, which

 started out as means of exchanging non-commercial information

 between government and academic research centres, has reached a

 critical phase in its history: it has become so large, and is

 growing so fast, that the pressure for its full commercialisation

 has become intense.

 

 Between half and three quarters of all traffic on the network

 originates from commercial organisations. At present, these

 organisations are enjoying the benefits of a powerful, low cost

 and reliable international resource. But, like newcomers crowding

 onto an unspoilt beach, their activities may need to be controlled 

 - and they must understand the limitations - if the Internet is to

 reach its full potential.

 

 Recently, the US government funded research agencies which played

 such an important part in the establishment of the Internet

 concluded that funding or attempting to police huge parts of the

 Internet is no longer desirable or feasible. They have relaxed the

 rules on its use for overt commercial gain, and told individual

 research centres to fund their own Internet activities.

 

 With "the prohibition period over", says CIX's executive director

 Bill Washburn, the Internet is undergoing massive

 commercialisation. Driven by the prospect of a huge market, aided

 by new technology which makes it easier to use (see below), and

 encouraged by the new business culture, dozens if not hundreds of

 individuals, companies and consortia have started projects,

 businesses and technical initiatives to exploit the Internet for

 business.

 

 "We're talking about a whole array of business opportunities for

 conducting business in a new, high speed, heterogeneous

 environment of the future", says Washburn. The Internet, he says,

 can help companies reach untapped markets; overcome geographical

 constraints and language and legal barriers; sell more without

 fewer sales staff; and improve the speed of business

 communications and administration. It has the potential to

 completely revolutionise some areas of business.

 

 Tony Rutkowski, executive director of the Internet Society,

 agrees. "The Internet is now overwhelmingly commercial. There are

 an enormous number of commercial users". But the optimism should

 not distract from many of the limitations of the Internet. Even

 Washburn, an avid supporter of business on the Internet, counsels

 that commercials organisations would not expect too much. "For

 business, the first question should be 'How can I participate in

 the Internet and save money. Lets forego the question of how to

 make money for a while".

 

 There are several reasons why he gives this advice. The Internet

 was not designed to carry commercial traffic, and is fundamentally

 unsuited to it. Although it is being changed, this is happening

 only gradually and not without disagreements. There are several

 area where business practices need to be thrashed out and the

 technology improved if the Internet is to live up to its

 potential.

 

 According to Internet statisticians, the Internet is growing so

 fast it is almost out of control. Some twenty five million people

 are said to be on the Internet, and every twenty minutes a new

 user is connected. At current growth rates, says Tony Rutkowski 

 of the Internet Society, the number of registered computers on the

 Internet will surpass the number of people (five billion) on the

 planet by the year 2001. With figures like these, it is not

 surprising that businesses are rushing forward to participate. But

 they do not tell the full story. The size of the business audience

 is far smaller.

 

 John Quarterman, president of Texas Internet Consulting and the

 publisher of Matrix News, points out that measuring the size of

 the Internet is a very complex problem: " No-one knows its true

 size". His research suggests that while the Internet is

 undoubtedly the largest network in the world, and is growing fast,

 its size is often exaggerated.  The most widely used method of

 measuring the size of the Internet is to use the known number of

 IP (internet protocol) hosts - these are the computers that have

 an internet address. There were about two million of these in

 December 1993, according to Quarterman. Some estimates argue that

 for every IP address, there are ten computers or end-user

 workstations which can connect to each Internet node. This makes

 20 million Internet users,  a figure that is frequently rounded up

 to 25 million to take account of recent growth. A more realistic

 estimate, used by Quartermann, is 7.5 users per host, making 15

 million Internet users.  Quartermann also points out that many of

 these Internet hosts, for a variety of reasons, are not actually

 reachable; in fact, he says, the 2 million figure of IP hosts may

 only be 0.5 million. This means the number of internet users could

 be as low as 3.75 million. In all likelihood, the number of

 Internet users is nearer 15 million than 3.75m - still an enormous

 number but short of the some of the more dramatic estimates for

 the Internet. These figures do not, however, include people who

 can access the Internet via commercial gateways, such as MCI Mail,

 Sprintmail or CompuServe. The availability of Internet services

 across gateways is usually restricted to electronic mail and some

 access to newsgroups.

 

 The little demographic research that has been undertaken (again,

 by Quarterman of Matrix) suggests that most users are young (below

 25), and nearly three quarters are either students or are working

 for educational or government establishments. 

 From a business point of view, this a difficult demographic

 profile, since while the number of budget holding decision makers

 may still be high, they will be very dispersed. Its promise lies

 not in today, but tomorrow, when many of today's Internet users

 more responsible positions. The problem is further compounded

 because there are no centralised directories and the sheer size of

 the network makes targeted communications extremely difficult.

 There are already signs of a market developing in lists of

 newsgroups; this may later extend to mail addresses of influential

 users.

 

 If the demography of the Internet is a problem for organisations

 new to the Internet, overcoming the sentiment against its

 development for business is likely to be even more so. The

 reaction against Canter & Siegel is a clear demonstration of what

 happens if an organisation oversteps the mark. "For purists, the

 use of the Internet for commercial activity is still pretty

 unacceptable", says Washburn of CIX.

 

 The lack of public email directories makes targeting difficult and

 advertising and promotion much more desirable; in theory, the way

 to do this is to post electronic promotional messages to the

 Newsgroups, yet 'netiquette' forbids this unless done very subtly.

 This is one area where Canter & Siegel believes things must

 change. "It is highly presumptuous of those individuals to think

 that the Internet should not be carry advertising". says Martha

 Siegel. She is now forming a company to help people do exactly

 this, a move which will certain bring her into bitter dispute with

 many Internet users.

  

 According to Bill Washburn of CIX, advertising is not expressly

 forbidden - but 'in practice there are real constraints'. Those

 that do advertise, he says, "must not be intrusive, especially in

 a junk mail sense. You need to provide value and not send out what

 looks like, feels like or sounds like propaganda".

 

 Mike Godwin of the Electronic Frontier Foundation, a consulting

 company, agrees: "There have always been people who try to make

 money [over the Internet]. Don't confuse advertising with

 obnoxious and intrusive advertising".

 

 At present, the battle between those pushing for more

 commercialisation, and those who want less, is finely balanced.

 But as pressure grows, the temptation to push back the barriers of

 netiquette will grow, provoking angrier reactions from the groups

 against commercialisation.

 

 The impending battles have no adjudicators, since the Internet is

 not under anyone's control. In theory, the battles could end up in

 the courts, with a radical liberal wing arguing against constraint

 of trade and alleging nuisance against those who try to prevent

 them sending out messages; against them, network access providers

 could claim that breach of contract led them to cut off Internet

 access.

 

 There have been occasional calls for bodies such as the standards

 setting IETF (Internet Engineering Task Force), the Internet

 Society or the CIX (which represents commercial network providers)

 to take responsibility and try to put forward some kind of

 constitution which could be written into network access providers

 contracts. However, none of these bodies is likely to feel

 remotely prepared for such as massive policing job.

 

 Washburn is hoping that some kind of community spirit prevails.

 "There is a huge reservoir of sentiment for keeping the Internet

 viable. I would say lets puts the community on top of the priority

 list and work out what we can do in terms of sentiment".

 

 If this spirit does not prevail, the newsgroups could become

 swamped with messages, forcing details of their existence to be

 kept quiet or even password protected. Similarly, many electronic

 mail addresses will be forced to go "ex-directory". Both measures

 would harm the openness, and to some extent the usefulness, of the

 Internet.

 

 The Internet is practically free to most users. However, the

 growing numbers of people using the Internet, largely fuelled by

 the development of the world wide web and the Mosaic graphical

 interface, is starting to cause problems. Several providers of

 information have been swamped by demand, causing them to re-assess

 whether, or indeed why, they should invest to maintain their

 levels of service to non-paying customers.

 "In my opinion, things must change. A lot of people have put up

 free information on the Internet because it was mutually

 advantageous. Now, some things are so valuable and so popular that

 the suppliers can't afford to cope with the level of demand" says

 Washburn. As a result, some academic sites are turning over their

 resources to companies who charge commercial users but not

 academic ones who provide complementary information.

 Jon Crowcroft of the University College and an Internet committee

 member, agrees. "Commercial users may get charged where academic

 ones do not" he says.

 Such restrictions are unlikely to slow the growth of the Internet,

 but they will diminish its appeal. The likely impact is that

 information and services on the Internet will start to fragment

 into categories: some will be free but have little commercial

 appeal; some will be free but will be 'sponsored' by advertisers;

 and some will be restricted or charged by connection time or a

 'licence fee'.

 

 In the information and publishing business, commercial providers

 are already starting to develop strategies around this fragmented

 model. The London Financial Times and Reuters, for example, are

 both evaluating plans to offer some services over the Internet

 using passwords to control access. Uses will then be charged by

 connection time. Internet Publishing, a London based electronic

 publisher, takes a different approach. "Our philosophy is that

 there should be no barriers on entry. Those that put up boundaries

 will quickly be ignored" says Eamon Wilmott.  Internet Publishing

 publishes valuable  newsletters paid by sponsorship, and will

 offer books and software using a model similar to the OMG 

 

 Business applications demand security, especially if money is

 involved; the Internet has a dire reputation for viruses, hacking

 (unauthorised entry into computers) and the theft of passwords. A

 recent example occurred in New York when a network access provider

 discovered that someone had been "trapping" passwords and email

 addresses. It never discovered who or why.

 

 The security problem is non insurmountable. As Dr John Crowcroft

 points out, "the attitude of people on the Internet is not

 immature. They understand that security is the responsibility of

 the end system".

 

 There are several ways to do this, starting with "firewalls".

 These ensure that messages from outside the Internet cannot get

 beyond the Internet gateway and into a company's main network. The

 result, however, can be to restrict use of the Internet among

 employees to simpler services such as electronic mail. Moreover,

 the Increasing use of passwords by all users is likely to inhibit

 the free access that runs deep in the Internet culture.

 

 One encryption method is now becoming established as standard - a

 'dual key' system known as "Pretty Good Privacy" or PGP. PGP

 appears to be uncrackable, and is the subject of enormous

 controversy, partly because of a dispute over the ownership of

 algorithms, and partly because the US government - and some others

 around the world - do not want it to be used commercially. The US

 government is pursuing a case one of PGP's developers, Philip

 Zimmerman, alleging that he put PGP onto the Internet, thereby

 violating export controls.

 

 Is PGP going to be secure and flexible enough for business use? It

 is not clear. Sandy Whitsun, Electronic Data Interchange business

 manager for Hewlett Packard has been grappling with the problem:

 "The question is "What is security? What methods do you use? Is

 encryption enough? If so, which method do you use?" She believes

 there are different answers to these questions depending on the

 application. A multi-level approach might become common on the

 Internet. "There are some transactions we couldn't care about who

 sees them or when they get there", she notes. For others,

 encryption would be a minimum requirement.

 

 A partial solution would be to use the X.400 message handling

 standard developed for business purposes by the International

 Standards Organisation. This is already the chosen method being

 used by thousands of companies representing millions of users who

 interlink their electronic mail and other systems. They are

 unlikely to move such sensitive applications onto the Internet

 until there is a clear business case, and security is guaranteed,

 says Roger Dean of the European Electronic Messaging Association

 (EEMA).

 

 X.400 standards can be used over the Internet, and many

 multinationals already do. But is the security of  X.400 preserved

 when an Internet backbone is used? "That isn't clear" says Dean.

 

 There is a further problem: it is extremely difficult to  prevent

 surreptitious traffic analysis over the Internet. This can matter

 in some situations: for example, two companies that suddenly start

 sending large volumes of information to each other could arose the

 interest of takeover speculators. The solution is for the network

 access providers to agree to continually generate random messages

 as background noise. No such agreement looks likely for some time.

 

 "We need a bank to put funds transfer onto the network. Whoever

 sets something up quickly and deploys it should win the spoils",

 says John Dawe, head of Internet access provider Pipex.

 

 To date, there are at least three initiatives aimed at putting

 some form of funds transfer onto the Internet. They include

 Commercenet in California ; the Internet Engineering Task Force

 using proposals from Bell Arbacus; and Digicash in Holland, which

 is developing a system using electronic tokens.

 

 None of these systems is yet up and running, and the transfer of

 money securely remains a major problem. Today, the most common way

 to send money is by sending a credit card number in an electronic

 mail message - a method which is clearly vulnerable, since mail

 messages frequently get intercepted, copied and distributed

 without the sender knowing.

 

 In the absence of a better method, groups such as Commercenet in

 the US are likely to use credit card numbers along with

 encryption. However, this is only suitable for small sums, not for

 large or bulk payments.

 

 In Silicon Valley, California, home of many leading computer and

 router companies, the Internet culture already runs deep. But its

 use for business purposes has, with a few notable exceptions,

 largely been restricted to electronic mail and the exchange of

 non-commercial software.

 

 The CommerceNet initiative aims to change all that. "This is

 monopoly busting. We think there will be an immediate demonstrable

 edge in competitiveness for companies that are participating",

 said Jay Tenenbaum, chairman of Enterprise Information

 Technologies when the plan was announced earlier this year.

 

 The lead partner in Commercenet is EIT, but it includes major

 computer companies such as IBM, Hewlett Packard and Digital

 Equipment, component suppliers such as Intel and National

 Semiconductor, and, so far, one bank, the Bank of America. The US

 Government's Technology Reinvestment Program is providing $6

 million.



 CommerceNet sets out overcome practically all the shortcomings

 that affect the Internet today. It will develop secure multi-media

 messaging, build up privately and publicly accessible mail

 directories, and is exploring the best methods to support

 electronic funds transfer. It will use the PGP (Pretty Good

 Privacy) encryption technique, allowing for the transmission of

 electronic letters of credit and digital electronic signatures.

 Special applications, involving electronic data interchange, are

 being developing for the electronics industries in California.

 

 Participants in the group are developing a start-up kit for easy

 access, which will include the supply of high quality affordable

 connections. Commercenet  be heavily promoted - but only in

 Northern California. In a marked switch away from the global

 emphasis of almost everything on the Internet, Commercenet is

 designed to serve a local geographic area. Other similar

 initiatives are expected to be developed in the Boston, Austin

 (Texas) and Champaign-Urbana, Illinois.

 

 In spite of the commercial opportunities, the goal of Commercenet,

 is more about saving money than making it, says Sandy Whitson,

 electronic data interchange business manager for Hewlett Packard.

 "Our motive is the proliferation of the technology - its part of

 the evolution of electronics".

 

 Tony Rutkowski of the Internet Society thinks the lack of a funds

 transfer system will not inhibit the Internet's appeal. "There are

 an enormous number of uses that don't require an exchange of

 money" he says, citing research and development, public relations,

 customer support and distribution as functions which can all

 benefit from using the Internet.

 

 A common complaint about the Internet - that it is too difficult

 to use for most business people - is now rarely heard. The reason

 is the development of graphical navigator's such as Mosaic and

 Cello, and the multi-media World Wide Web These systems have

 transformed the Internet, making it far easier to use and

 attracting new types of user and boosting traffic levels.

 "Although the Internet is huge, it hasn't really started to

 happen" says Eammon Wilmott of Internet Publishing. He sees the

 Web as the "first superhighway application" which will fuel

 spectacular growth. "A year or two ago, you had to put in twenty

 hours to learn to use Internet. Now its point and click".

 

 The effect of Mosaic has been dramatic. From a standing start, the

 World Wide Web accounts for 10% of Internet traffic, and is rising

 fast. The National Center for Supercomputer Applications in

 Illinois, which developed the Mosaic application, is getting 1.3

 million connections per week, simply because Mosaic 'defaults' to 

 logging into its computer unless it is changed by the user.

 

 The Mosaic software, which runs on Windows, X-Windows  and the

 Apple Macintosh, is free. However, several software suppliers are

 moving into the market, buying rights from the NCSA to develop

 more advanced applications around the product. These include

 Quarterdeck, Santa Cruz operation, and several smaller companies. 

 

 If Mosaic has put strain on the network by boosting traffic

 levels, the plans of the major software companies could put some

 sites into crisis. Microsoft, Novell and Lotus are all developing

 seamless interfaces for linking their PC and network software into

 the Internet. This is certain to encourage a new wave of end users

 to join the Internet.

 

 The traffic levels are certain to create problems, although Tony

 Rutkowski points out that they will be most significant at the

 periphery. "The big carriers have all the bandwidth in the world"

 he says.

 

 Because of the impact of the world wide web, slow response times

 and clogged telecom lines are already common at some popular

 sites. An example was seen at the Winter Olympics, where the

 information service was overwhelmed with 100,000 enquires on the

 first day.

 

 There are three parts to this problem. First, many sites will have

 to buy more or larger computers to support a growing number of

 log-ons (and to support a fast growing routing table of at least

 10 megabytes); second, many users will be under pressure to

 upgrade their telecommunications lines; and third, the Internet

 addressing will ultimately have to be changed.

 

 The first two problems are financial, and will be dealt with on

 that basis. Those that cannot afford to upgrade their computers or

 lines will have to restrict access. The third is technical. "The

 original design of the protocols had not envisioned this kind of

 success. The routing of the network is becoming unmanageable" says

 Jon Crowcroft, a member of the IETF's next generation working

 group. "We have until 2008, plus or minus three years, until we

 run out addresses" 

 

 The IETF is considering the introduction of a new protocol, which

 has more hierarchical levels and longer addresses. But it will

 have to be compatible with existing systems, which cannot be

 changed. Indeed, it is not even clear that the IETF would have the

 power to force through changes if Internet users did not want

 them.

 

 Some believe that the introduction of a new addressing system -

 which would necessitate changes to software - will further

 fragment the Internet into different groups with different

 technologies and profiles. 

 

 THE SUPERHIGHWAY MYTH

 The dramatic take-off of the Internet has co-incided with another

 related development in the field of computing and

 telecommunications: the first experiments in video-on-demand and

 the first plans to build high-bandwidth information superhighways.

 It inevitably leads to the question: is the Internet already

 established as *the*  international superhighway.

 

 The question can be broken down into three: will the Internet ever

 be able to support the kinds of high-bandwidth interactive

 applications which are associated with the idea of a superhighway;

 is it commercially suited to carry such applications; and is there

 or will there ever be such thing as a superhighway.

 

 The Internet's ability to support high-bandwidth interactive

 applications is being explored by the IETF at present. Time

 Warner, the entertainment giant, and Reuters, the information

 services company, are among those that are believed to have

 submitted ideas. However, it seems unlikely that existing lines or

 protocols will easily support the piggy-backing of continuous

 video streams; any video of demand type applications are likely to

 be on the periphery of the Internet among organised groups of

 users.

 

 The IETF is also looking at other issues which the Internet needs

 to address if it is to support large numbers of business and

 mass-market applications. These include support for much larger

 numbers of devices, mobile computing, guaranteed response times,

 multicasting (simultaneously sending streams out to many stations

 at once) and media independence (ie support for satellite, radio,

 dial-up lines etc). 

 Some of these technologies are relatively simple to fix, some will

 take many years and some may require the use of non-Internet type

 protocols. Even Internet enthusiasts think it unlikely that the

 business community will wait until the Internet technology is in

 place before pushing ahead. "It is easy for people that are coming

 from one point of view - telecommunications, of Cable TV, or the

 Internet, to say that they have the foundation of the

 superhighway. My view is that it will be a motley combination of a

 few" says Bill Washburn of CIX.

 

 Has the use of the Internet been over-exaggerated. Certainly, the

 difficulties have been played down and the size, scope and

 uniformity of the Internet has been played up. As it is currently

 set up, the Internet is a difficult commercial environment for

 many if not most types of business, especially those that need

 guaranteed levels of security, guaranteed responses, the ability

 to mass market services and impose charges.

 

 Its evolution as a business tool will happen more slowly than the

 growth of traffic levels suggests. In the process, business users

 will have to deal with the increasing fragmentation of the

 network.

 

 Two of the Internet's great advantages are in danger of eroding:

 Its homogeny, in that most users today have similar attitudes,

 similar access methods, similar technical requirements and use

 similar protocols; and its ability to provide free access to

 virtually any host on the network.

 

 In its place, a larger, more powerful, but fragmented Internet

 will emerge. This will ultimately be more useful for business, but

 creative approaches will be needed to gain competitive advantage.

 In effect, the Internet will have ceased to exist as one entity,

 and become what it was always intended to be: a meeting point for

 thousands of smaller and often very different networks. 

 (C) Computer Business Review