1. Table of Contents

• 2. Executive Summary
  • 2.a. Purpose
  • 2.b. Why important?
  • 2.c. Major technological risks involved in this proposal
  • 2.d. Essence of the Research Plan
  • 2.e. Strategy for commercializing product
• 3. Research and Development Program
  • 3.a. Scientific and Technical Goals
  • 3.b Technical Risk and Feasibility
  • 3.c. Analysis of Technical Plan
  • 3.d. System Integration and Concurrent Engineering
  • 3.e. Broad impacts on US Technology and Knowledge base
• 4. Experience and Qualifications
• 5. Broad Based Economic Benefits
• 6. Plans for Commercialization
  • 6.a. Elements of the Commercialization Plan
  • 6.b. Technology Diffusion & Protection of Intellectual Property
• 7. Proposer's Level of Commitment & Org Structures
• 8. Budget
• 9. Bibliography

2. Executive Summary

The Coalition for Electronic Markets (CEM) proposes to address the business goals of this program by building, deploying, and operating as an ongoing commercial concern, an invocation-based revenue collection and distribution infrastructure for computer-based electronic goods. This would:

• support productivity of software users by allowing them to acquire unfamiliar software at negligible cost, while collecting revenue for the provider in proportion to the software's measured utility to the user.
• broaden software markets by lowering the threshold to market entry for producers and consumers regardless of company size or product granularity. The uniformly multilevel approach proposed here provides the same revenue collection infrastructure to ordinary citizens who might wish to package their tacit knowledge for sale that it provides to large software-producing institutions like Microsoft or Taligent and electronic document producing institutions like Addison Wesley or the Washington Post.

2.a. Purpose

We understand `software components', as used in the ATP solicitation, to mean objects smaller than computer applications. Our proposal addresses this objective by engaging the more general problem of multi-granular goods; goods that contain smaller-granularity goods inside as in Figure 1. We trust that an infrastructure that could support the emergence of a true information age economy for large and small producers and consumers alike is consistent with this solicitation's intent. We seek a multigranular structure of production for electronic goods that is comparable to the vast economic tree that provides tangible goods such as automobiles and food products. This distinguishes CEM's approach from the single-granularity approaches used already (license servers, copy protection technologies, shareware, subscription fees, downloading fees, etc).

Commerce has been multigranular since antiquity. Terms like `organized human society' refer to the specialization of labor that distinguishes modern society from the unspecialized hunter-gatherer society of our ancestors. But paradoxically, our most advanced industry, software engineering, is the last refuge of stone age human organization. Instead of assembling prefabricated components from a stockroom supplied by others, every software developer fabricates each component from first principles.

This is not due to lack of desire or recognition of what is widely known as the software crisis. The problem is much more fundamental. There is no stockroom of components to choose from due to the lack of a robust incentive structure to persuade people to provide them.

Physical laws, such as conservation of mass, make tangible goods hard to copy. This is the basis upon which commercial exchange transactions rests. It is easier to buy copies of goods than to replicate an existing copy. This understanding breaks down for easily-copied goods such as computer software

This proposal will call the commercial incentive that keeps manufacturer's stockrooms filled pay-to-acquire . Pay-to-acquire relies on physical conservation law to supports commerce in multigranular goods. For example, the pay-to-acquire basis for commerce in automobiles also supports commerce in the spark plugs, engines and tires from which automobiles are comprised. Every level of the structure of production understand that they'll prosper in proportion to their contribution to the prosperity of those who use the product of their labor to build higher-granularity goods. We engage in mining, or in wheat or dairy farming, or indeed in any agricultural or manufacturing activity, because we understand that physical laws of conservation will effectively couple our prosperity with the well-being we contribute to the economic tree that relies on our products.

Although pay-to-acquire fueled the economic engine that brought the achievements of the Age of Manufacturing, it does not work for the easily copied goods of the Age of Information. Examples of such goods are electronic documents, software applications, and the reusable software components from which applications are composed; the ephemeral goods of the Information Age. It doesn't work because there is no way to incent potential providers of software components to provide them Farmers understands that each sale by the baker will be guaranteed (by physical law) to renew the demand for wheat. Software component vendors know that their goods can and will be replicated without compensation whenever new copies of application based on them are sold.

Figure 1: The proposed revenue infrastructure handles the flow of revenue from consumers to the appropriate members of a multigranularity structure of production for electronic goods.

2.b. Why important?

The breakdown tacit understanding that made every level of the economic tree work as a unit is so fundamental that it is rarely articulated explicitly. Its breakdown for the easily copied goods of the information age can be observed in the following problems, all of which can be understood as diverse symptoms of the ownership crisis resulting from the advent of easily copied goods such as software:

• Software Piracy: How to control revenue losses from unauthorized copying of software, not just within the U.S., but also after it has been exported to countries where copying of software is not illegal, nor in some cases, even regarded as immoral?
• Electronic Publishing: How can book, magazine and newspaper publishers exploit electronic networks to reduce their present reliance on paper-based publishing, and the consequent degradation of the environment from deforestation, pollution from pulpwood processing, and trash disposal?
• Reusable Software Components: Why is most computer software fabricated from first principles instead of being assembled from reusable software components?
• Software Crisis: Why has software productivity only grown arithmetically during the very period that hardware productivity has grown exponentially?
• Dual-use Software: How can we control how dual-use software is used after it has been exported, for example, during wartime?
• Mass market for computer hardware: The quality and abundance of goods within existing networks like Internet, Compuserve, Prodigy, is not sufficient to persuade a mass market to buy the computer and networking infrastructure for accessing it. Quality content is more abundantly available in newsstands, bookstores and other non-electronic media. The market for hardware is limited by insufficient content, not carrier.

2.c. Major technological risks involved in this proposal

CEM's proposal involves building and deploying a system that integrates three well-developed components - computers, telecommunications, and financial institutions - to produce a business enterprise analogous to a credit card company.

This involves both technical and business risks. CEM regards the technical risks as straightforward compared to the business risks involved in deploying a paradigm shifting but financially sensitive system. This proposal does not dwell on technical issues such as user authentication, performance, and the challenges of implementing a large distributed system. We hope to use existing technology insofar as technology already exists. We will rely on technical means in combination with social and legal sanctions, restricting risks by limiting initial deployments to markets where risk of fraud is acceptably low.

These risks arise from the likelihood that the revenue collection infrastructure will involve at least some software during initial infrastructure deployments. This is to avoid having to provide hardware updates within every component-using computer. In the long term, these risks can be surmounted by building the infrastructure within tamper-resistant silicon and by guarding critical hardware-software interfaces with encryption. In the short run we'll proceed by deploying initial instantiations to markets where the risk of tampering is acceptably low. Such a market is to established companies who wish to provide software to their employees. Fortunately, this happens to be a market that is commercially significant today, and which is already impatient with the revenue collection alternatives that are presently available to them.

Secondary risks arise from the separation of invocation counting from financial charging resulting from the need to avoid having the revenue infrastructure dictate or restrict terms component owners may offer to their customers. This risk will be protected by enabling end-users to observe charges to their accounts after trial transactions, and by enforcing sanctions against vendors found guilty of fraudulent charging.

2.d. Essence of the Research Plan

Insofar as possible, CEM does not expect to engage in basic research to advance the state of the art in computing, communications, encryption or silicon fabrication. CEM's function is to deploy existing technologies such as cryptography and tamper-resistant silicon to address a long-standing commercial need. However there's another meaning of the term `research' in which this proposal is ground-breaking indeed. Although we expect to avoid techno-centric research, we'll be fully involved in research in the human-centric sense. We'll be pioneering a new way for people to cooperate, coordinate, communicate and compete as members of an global information age community.

2.e. Strategy for commercializing product

We plan a three-phase roll-out. The first phase, which is already inplan by at least two coalition members, will deploy existing license server technology. ATP funding for this phase is not needed. The second stage, which will probably involve a software-intensive infrastructure, will be deployed to the same market segment as the previous one, and will thus benefit from the market presence then established. The third phase, which will probably need to be hardware-intensive, will follow as dictated by the risk of fraudulent behavior we perceive as deployment proceeds to broader markets.

Due to the multigranular nature of our approach, CEM doesn't view the market in the traditional way; as component buyers versus sellers, since every seller is also a buyer of the components used in building the product. Even ordinary end-users, the traditional buyers of today can have the incentive to package their tacit knowledge as software components or electronic documents for sale. Nonetheless, the initial market is stratified in the traditional manner and we'll address it on those terms:

• End-users: The benefit to end users is free access to software with low transaction-cost payments (once a month billing), lower prices (driven low by competitive pressures), and flexible terms and conditions. We expect that flat rate pricing (pay to own) will be preferred by some customers. Others will prefer variable pricing in which charges are precisely related to measured utility. Both are supported by our approach.
• Application vendors: The selling point to vendors is the ability to distribute software in an extremely low-cost fashion (via computer networks) while robustly protecting their revenue stream, in both cases compared to what they're using today. This author can attest from personal experience that the costs of software are not the salaries of the programmers that built it but in advertising, marketing and distribution. By allowing software to "advertise itself" via networked distribution, and by limiting todays revenue loss to piracy, the dominant cost are reduced to the point that prices can be lowered, thus broadening the market.

CEM's coalition building activities have not matured to the point that we can propose marketing and sales staffing for implementing this strategy. This will be remedied depending on the outcome of this proposal.

3. Research and Development Program

CEM's proposal addresses the socioeconomic question of how to couple the prosperity of those who produce fine-granularity electronic goods with the contribution they bring to the well-being of those who benefit from their labor. This section will state the problem in socioeconomic, i.e. human-centric instead of techno-centric, terms. Subsequent sections will focus on the technical implications of building a widely deployable implementation.

3.a. Scientific and Technical Goals

CEM's goal is to deploy a revenue collection infrastructure that can provide a basis for multigranular commercial exchange. By this we mean incentives analogous to those that apply in tangible domains. The emergent integration of computing and communication technology of the Nationwide Information Infrastructure provides a way to supply software invocation information to financial institutions for billing. With the aid of the revenue collection infrastructure outlined here, this could provide a revenue channel capable of supporting a multigranular information age economy in electronic goods.

Contrast a farmer who grows wheat with a software developer who builds reusable software components. The farmer is guaranteed by natural law (conservation of mass) that each sale by the baker will deplete the supply of wheat and thus replenish the market for future wheat sales. This yields an incentive structure that mobilizes every lower level in the structure of production to pull in the same direction, each contributing capital, labor and knowledge to help the baker sell more bread. Since everyone understands that wheat is not copyable, the farmer and the miller are both assured that they will prosper as the population grows, as wheat products become more popular, and as people find new ways to use wheat. Although this principle is so basic that it is rarely articulated explicitly, this yields the economic engine that supports what we mean when we speak of "organized" or "advanced" human society.

Those who might produce electronic goods do not enjoy these advantages. When a supercomponent builder, say Microsoft, sells an additional copy of a word processor, the market for word processor subcomponents is not replenished. Any subcomponents within the word processor will be automatically copied without compensation. Since no per-copy revenue flows to the sub component owners, no software components `industry' has arisen that is comparable to what is common in manufacturing domains.

The problem originates from deep causes. Natural laws such as conservation of mass apply only to goods made of atoms. They do not apply to electronic goods made of bits. Thus electronic goods providers encounter an ownership crisis to which they can turn to neither natural or social law for a resolution. But at the same time society is forced to seek a solution by the forces that are driving us towards a global information age. This is the challenge that CEM's proposal addresses.

3.a.1 Outline of Technical Approach

CEM's approach originates in the following technical observation. Software is unable to monitor its acquisition but trivially able to monitor its invocation. It is easy to make software count how many times it has been invoked, but hard to make it count how many times it has been copied. Rather than continually fighting this characteristic with legal/moral restrictions and copy protection technologies, CEM proposes to build on it as the solution.

We propose a revenue collection infrastructure in which revenue collection is based on the invocation of software within the end-users' computer. This makes revenue collection independent of how software copies were acquired. Since each copy used means another paying customer, copy protection is neither needed nor desired. Vendors are incented to give the bits away free in expectation of invocation-based fees. They know that their software will use the revenue collection infrastructure as a subroutine to convey invocation information to a financial institution for billing and financial settlements. The supporting infrastructure is shown in Figure 1.

Figure 1: The integration of computers and communications technology with financial institutions creates the possibility of an invocation-based market mechanisms, where revenue is collected when electronic goods are invoked as distinct from when replicas are acquired. This is a significant development because this could provide the basis for a solution for the ownership issues that are created by goods that can be copied without cost or loss and transported at literally the speed of light.

This introduction has stated the problem in the most general possible terms in order to allow for granularities of goods that are not commercially viable in the absence of the proposed infrastructure. The remainder of this proposal will concentrate on granularity levels that are commercially significant today since this will be where initial deployment will be targeted.

Revenue collection mechanisms for software components must be engineered and implemented in light of the opportunities and challenges presented by fine grained components as compared to today's large horizontal applications. These requirements include: ubiquitous availability and usage by independent software vendors, operating system providers, and financial institutions; Ease to use by component developers and users; resistance to fraud; protection of the privacy of invocation information; inexpensive to administer and to use, with a minimum of additional path length within the user's computer for recording usage; legal, where all parties have sufficient safeguards; and provide sufficient opportunity for developers and users to review and understand their obligations and afford them the opportunity to review usage information at any time. Figure 2 shows the proposed solution in greater detail. Each developer uses a development environment to develop and test components of any granularity. By interacting with the Clearing House, the developer registers the product to acquire a product identification number. The developer embeds this number in the component to identify it, and thereby its owner, during subsequent invocations of the component by its users.

Although the Clearing House might choose to involve itself in advertising and distributing registered components, the revenue collection mechanism is in no way dependent on how the components are distributed. For example, users could acquire software from download areas on networks, acquire copies from friends, or from free copies mailed them by vendors. Vendors are incented to provide free copies because the components are non-functional except by registered infrastructure users. Users register with the Clearing House in order to establish an account, whereupon they are provided the infrastructure components labeled interface and cache in Figure 2. They're also issued an identifying key, which is analogous to a credit card number in that it identifies a particular account-holder. Most likely this key will be the private half of a public key pair, and would be supplied in a device such as Dallas Semiconductor's Button^reg. technology[1].

Figure 2: The proposed revenue collection in infrastructure.

The user inserts the button in a reader and proceeds to use software When the top-level application is run, it, along with each of its sub components, invokes a pair of functions in the infrastructures API. Each component's owner uses these to bracket each section of value-bearing code. The first function, query, confirms that the revenue collection is present and that the user's account balance is within threshold. The second function, commit, records that a value-bearing segment has been invoked and that the invocation is to be recorded as the basis for subsequent billing.

The time sequence in which these two instructions are invoked by a nested hierarchy of multigranular components generates the invocation data upon which invocation-based billing is based. Each instruction records the fact that it has been invoked by appending its name and arguments to a list maintained by the cache in Figure 2. A mandatory argument of both functions is the id of the component being invoked, or in the case of a top-level component the id of the user. The functions also accept optional arguments that component owners can use to record such information as the date and time in case they chose to provide time dependent terms and conditions.

Figure 4 shows an example of the invocation record. On a flexible schedule (probably monthly) or when the user's computer is connected to suitable communications, the component usage information is uploaded to the Clearing House. After the on-line financial transactions are processed (see below) and debit and credit processing have concluded, the Clearing House signals the PC-resident infrastructure to reset itself and thus allow subsequent revenue cycles to proceed.

Figure 4: As the component hierarchy on the left is executed by the computer (top-left to bottom-right order in this picture), the query/commit instructions at the beginning/end of each component generate the invocation record at the right of this figure.

Invocation records are converted to financial amounts at the Clearing House. That is, the infrastructure components in the end-user's PC deals entirely with invocation counts. This is a crucial design feature. It means that the infrastructure need not dictate or restrict the terms and conditions that component owners might impose on their goods. The infrastructure only supplies a general basis, invocation records. Component owners use this basis to support whatever terms/conditions they please. They do this by providing Terms and Condition Algorithms that the Clearing House can use to convert invocation counts to financial amounts due.

The charge computed by each owner's algorithm is, apart from rules of fair trade to be described in the next section, left entirely to the business discretion of the component owner. These algorithms can support traditional fixed rate (pay to own) charging as easily as they can radical incremental approaches, such as pay-per-use, in which charges are proportional to measured utility. For example, the pay-to-own terms and conditions that dominate the software market today can be provided by an algorithm that charges the pay-to-own fee the first time the algorithm encounters an invocation by a particular owner. Trial subscriptions, where the fee is incurred only after a period of trial use, involve delaying the charge by the advertised number of invocations or duration of time. Incremental, or pay-per-use, charging is also possible. The infrastructure does not require pay-per-use to be any more common in the future than it has been in the past, but it does not inhibit it either. The market is left free to select the products that offer the terms/conditions that the market prefers.

3.b Technical Risk and Feasibility

The technical risks of the mainline technical aspects of this project are not as severe as the social, i.e. non-technical, realities of a paradigm shifting project involving deployment of a financially sensitive system and protecting it from various forms of fraud. Therefore this section will not dwell on straightforward technical issues such as user authentication technologies, computer and network performance, and the challenges of implementing a large distributed system. We'll concentrate on the social challenges of protecting against fraud while insuring individual privacy.

The dominant risk arises from the fact that in invocation-based charging the fundamental commercial transaction, invocation as distinct from acquisition, occurs in the privacy of the user's home or office computer instead of a public space such as a store. Since early instantiations of the infrastructure may involve at least some software, the infrastructure is subject to tampering by those bent on fraud.

Technological means such as strong cryptography and tamper-resistant silicon, can in principle provide arbitrarily high protection. However CEM sees its role as commercial as distinct from research and development. CEM does not expect to engage in research to provide rigorously crack-proof technology to prevent fraud altogether. We expect to proceed as credit card companies do, in an iterative fashion that takes advantage of readily available technology while relying on social sanctions, such as law enforcement, to keep fraud to a level that is tolerable within the market conditions of the moment.

On the other hand, vendors wouldn't trust their revenue stream to invocation meters that could be erased by simply disconnecting their computer's power once a month. Nor would users accept a solution that uploads potentially sensitive invocation records to public networks as cleartext. This would create threats to individual privacy unheard of until now. We intend to rely on a rational combination of technical and social sanctions appropriate to the level of trust between buyers and sellers at each stage in the development of this market. This is distinct from indefinitely waiting until technology can guarantee rigorously fraud-proof systems.

In practice, we shall confine any early software-intensive incarnations of this system to markets in which the risks of fraudulent behavior are acceptably low. Fortunately these are the very markets that are financially significant today. The dominant producers today are large software companies, who have too much to lose to engage in overcharging customers or underpaying suppliers.

CEM is aware of two options that could accelerate availability of hardware solutions. Dallas Semiconductor is interested in security-related applications of their Computer In A Can(TM) and Button(TM) technologies. The second option is that Wave Systems (see section 6b) is developing silicon that may be suitable for this project. Their chief executive, Peter Sprague, has expressed willingness to participate in this coalition once the distractions of this summer's public stock offering are past.

3.c. Analysis of Technical Plan

This section examines the technical plan by responding to often-asked questions about this approach.

What prevents users from giving software away to friends? Nothing. Copying is actively encouraged. Basing revenue collection on invocation instead of acquisition means that each new copy brings another paying customer.

What prevents a user from running the software on an unconnected machine, refusing to cooperate with uploading, or refusing to pay the bill? The software will (politely) refuse further access if connectivity is refused or bills are not paid.

Can this work so long as not all computers are connected networks? Direct connections are not required. Usage information is held in a tamper-resistant cache and uploaded periodically (monthly). Sufficient connectivity can in principle be provided through unsophisticated technologies such as modems or even (far less satisfactorily) Smartcards or even floppy disks to be mailed back and forth in the mails.

Doesn't invocation-based charging run counter to consumer resistance to usage-dependent charging? Usage dependent charging is not implied. The charging model is determined by terms and conditions algorithms that component owners provide. The infrastructure only conveys usage counts, not monetary amounts, to the clearing house, where the terms and conditions algorithms convert them to monetary amounts due. These algorithms can support the pay-to-own terms and conditions that dominate the market today as easily as they can support radical incremental models in which revenue is proportional to the software's measured utility to the user.

Isn't software so malleable that the system can be subverted? Software is not unique in this respect. The perennial race between bank vault manufacturers and the safe cracking industry shows that even bank vaults are malleable in this sense. Robust commerce demonstrably occurs in shopping malls where the barriers to shoplifting are as negligible as the obstacles to piracy in computers.

What prevents hackers from a distributing a patch that disables the revenue collection infrastructure? In the long term, the infrastructure will be instantiated in tamper-resistant silicon. However in the short term, the software-based instantiations will be at risk. We expect to control this risk by confining software-based instantiations to markets where the risk of tampering is acceptable, such as large companies. We'll also rely on the social sanctions that presently inhibit large-scale copying of software. The license agreement for the infrastructure will identify tampering as fraud. Fraud would be dealt with as described in the terms of the license agreement that every infrastructure user would agree to before being assigned an account.

Could dishonest vendors advertise one price but provide TCAs that charges a higher one? Yes. This is another form of fraud. However it is easily detected, since every user is equipped to monitor their account balance, and can detect overcharging by observing changes after a trial transaction. Such vendors would be penalized by terminating the accounts of vendors found guilty of such practices.

Doesn't metering pose unacceptable threats to privacy? New systems brings new risks of abuse whose acceptability depends on the viewpoint of the observer. Millions of people chose credit cards over cash every day, giving up the privacy advantages of cash transactions in return for compensating advantages. On the other hand, CEM does recognizes that electronic privacy is an explosive political issue, and expect that this will be the single most critical factor in the market acceptance of our approach. Therefore, we intend to impose the strictest possible controls over invocation records, foregoing the advantages of even socially beneficial uses by guaranteeing that the data will only be used for billing.

3.d. System Integration and Concurrent Engineering

Although this proposal has emphasized invocation over acquisition as a basis for commerce in electronic goods, we don't think that invocation will replace acquisition entirely, neither in the short run nor, in fact, ever. As will be described in the commercialization section, we plan to begin operations with one of several candidate licensing servers now being used by CEM's member companies. This would provide a marketable service to today's software industry, provide a testbed and market for technologies being developed for the invocation-based approach, establish a commercial presence in the revenue collection market, and establish the synergistic relationship between renting and owning that is the norm in other industries.

Our concurrent engineering plan will be shaped and constrained by the commercial and technical interests of coalition members. They were specifically chosen for capabilities and interests that span the task components of this undertaking, from silicon fabrication, to computer software (object-oriented components and operating systems), to computer hardware technology, to communications technology, to financial institutions. Deployment will proceed in 3 stages.

• Stage 1: CEM will deploy existing acquisition-based license server technologies, as is already in-plan by at least two coalition members. Meanwhile development of acquisition-based technology will proceed in parallel, with Citicorp concentrating on interfacing the invocation-reporting mechanism with existing credit card operations, developing a front end server for running Terms/Conditions algorithms for translating invocation records into financial amounts due, and continuing efforts to support online connectivity between financial account management software such as Quicken. Other coalition partners will concentrate on downstream technologies at the end-users computer.
• Stage 2: Deployment of initial revenue collection infrastructure. The initial infrastructure may involve a minimum of special hardware, possibly confined to existing Button(TM) devices, which will be used for identifying and authenticating users. As a fall back plan, conventional (i.e. password-based) authentication is an option.
• Stage 3: The initial reliance on software will evolve to increasing levels of tamper resistance with greater reliance on silicon technology.

Should Wave Systems' plans materialize and computer vendors begin adopting a hardware-based ownership mechanism adequate to CEM's purposes, this could accelerate Stage 3 substantially.

3.e. Broad impacts on US Technology and Knowledge base

This topic will be addressed more completely in Section 5: Broad-based Economic Benefits. Briefly, the proposed system will enable the long-nascent potential of a reusable software components industry, which would make it possible for others to assemble software applications from off-the-shelf parts instead of by fabricating everything from first principles.

In the longer run, the proposal supplies an incentive mechanism that could mobilize the U.S. population to Information Age production, in the sense that traditional commercial mechanisms accomplish this for tangible goods. The same mechanism that could stimulate programmers to develop reusable software components could also prompt ordinary citizens to package their tacit knowledge in electronic forms for other citizens to use.

4. Experience and Qualifications

CEM's members were specifically selected for skill sets that span all the major elements of this proposal's infrastructure's technical, operational, social and organizational architecture:

• Banking, Financial
• Clearing House
• Infrastructure Software
• Infrastructure Hardware
• Software Components (electronic goods)
• Legal: coalition building, fraud prevention

By pooling the R&D efforts from CEM's members, we expect to design and prototype a revenue collection mechanism quickly. Each of the participating service providers has a vested interest in commercializing the resulting prototype. Although the implemented interfaces would be open and available to the industry for use, the participants in this joint venture will have several years of lead time in order to establish themselves before competition become effective.

4.a. Project Staffing

CEM members will cooperate in designing, implementing, deploying the overall architecture. System operational staff will be hired by the joint venture or subcontracted to external firms. Sections of the architecture will be assigned to each joint venture partner based on expertise. Interfaces between the assigned areas will be managed by a project lead. Work content and delivery dates will be committed and managed by each partner with overall project control by the project lead. Upon acceptance of this proposal, but before funding is issued, agreements between the partners will ensure the completion of assigned work and the continuation of the project in the presence of any partner's default.

4.a.1. Taligent (by David Sharp)

Taligent is an influential joint venture between Apple Computer, IBM and (recently) Hewlett-Packard. Its mission is bringing an advanced object-oriented operating system to market. Taligent's involvement in the coalition would facilitate early deployment of a software-only metering mechanism to producers and consumers within large corporations. As deployment extends to broader markets, their connections with dominant computer hardware vendors would facilitate the eventual migration of financially critical parts of the system from software to tamper-resistant hardware.

Taligent was formed to provide the operating environment for the next generation of desktop computing. It has sufficient expertise and resources to accomplish this task. Its preference is to develop the proposed revenue collection and distribution infrastructure in concert with other partners as this provides the opportunity for the broadest short term market acceptance. However, Taligent is also prepared to fund the development of a more narrowly defined revenue collection and distribution mechanism capable of addressing explicit requirements of its developer and user communities. Working in concert with other companies and with the assistance of ATP funding will broaden and enrich the scope of such efforts and ensure a more ubiquitous deployment.

4.a.2. Object Management Group (OMG) by John Slitz

OMG was founded in 1989 to develop a common architecture and interface specifications for distributed computing based on object technology. OMG currently represents over 400 companies and is considered a leader in the development of cooperative standards for Object Technology. While OMG participation in the Coalition will not guarantee adoption of the resulting infrastructure, as a OMG specification the flow of information on OMG activities can only be a benefit to OMG and the other Coalition partners.

A cardinal mission of OMG is the development of a market for object based products and services. For several years we have been studying the viability and operational concerns of a brokerage for software objects and information on all aspects of software development and usage. In 1993 the OMG copyrighted the name Information Brokerage (IB) for this project. The IB will create an electronic marketplace for information, technical and business, and then (or perhaps simultaneously) component level software. The IB will use the Object Management architecture and OMG Interface Definition Language as organizing guides for most component level software contained, viewed and sold through the brokerage. The IB will also list and sell software that does not conform to either of these standards. OMG believes that the central ingredient in a reuse (of commercial software component marketplace) strategy is a clearly articulated and widely accepted architecture supported by detailed, non proprietary interface specifications, as represented by the OMA and IDL standards.

OMG IB offers the Coalition an opportunity to use this program as a test center for the development of an invocation based revenue collection infrastructure. All users and providers of software would be offered the opportunity to use the invocation based system as a preferred method of making their products available. Because the OMG is a widely respected and active player in the emerging uses of object technology we see our participation in the Coalition as both helpful to us in meeting our goals of furthering a market for OT based products and as lending additional credibility to the coalition in the development of a widely recognized invocation system.

To pursue the Information Brokerage idea, OMG has entered into a partnership with Connect, Inc, a California based company. Connect will provide operational expertise, computer resources and widely available dial up access to the Information Brokerage. It is our plan to have the IB announced and operational by the end of 1994. Connect is a well established company recognized as a quality provider of commercial bulletin board services to a wide range of commercial companies.

4.a.3. Citibank

Dan Schutzer, Advanced Technology, and Frank Bamberger, a Citibank Vice President, are members of the Architecture Team for Citibank's Retail Bank. This team defines the services required to support interoperability on the NII, specifies how Citibank can take advantage of the opportunities presented by electronic commerce, and pursues other R&D opportunities.

Dan Schutzer anticipates that Citicorp's role within this project will be:

• To service the billing and payment function. Periodically, no more than once a day, receiving a batch of usage data originating from the software meters installed in the distributed software objects, and performing the function of processing the usage data against the agreed to terms of payment, billing and collecting payment from the using parties and paying and settling with the sellers and/or their banks. It is anticipated that the terms of payment would be separately negotiated in advance between the buying and selling parties, and the agreements certified by both parties and sent to Citibank for billing purposes.
• To assist in the commercialization plan, by helping to evaluate demand for the product, the associated operational risks, and identifying the roles and functions Citibank would like to play in a mature service offering, and developing a business plan and pricing strategy for the introduction of this service into the marketplace.
• Additionally, Citibank would be interested in evaluating the technology developed during the project, and assessing its viability for adoption by Citibank as building blocks for its offered services, should we decide to provide such services to the commercial market.

4.a.4. Dallas Semiconductor

DSC designs, manufactures, and markets a wide variety of electronic chips and chip-based subsystems. Many of their products are targeted at security and automatic identification applications envisioned in this work. For example, the DSC "Button(TM)" is a low-cost device that CEM expects to use for user identification and authentication. Other DSC products, such as their Computer In a Can(TM) product, could provide tamper-resistant storage for financially sensitive information such as the cache of monthly usage information.

4.a.5. Berkeley Productivity Group (BPG)

Berkeley Productivity Group (BPG) markets software products based on object-oriented technology to factories engaged in discrete manufacturing. Chris Lozinski provided the following text to include with this proposal:

BPG does business in California as a Sole Proprietorship. It has two product lines. The first is a horizontal development technology entitled "Windows Clients for OPENSTEP Servers." The second is the BPG BLOCKS MANUFACTURING CLASS LIBRARY.

The BPG BLOCKS library is the center piece of our participating in this proposal. However the DOME middleware, from Suite Software, which is part of our WIndows Clients for OPEN STEP servers solution, will also be part of this project. Thus our work on this project is closely tied to our central business directions, and leverages BPG's long-term investment in these technologies.

On this project, BPG will be represented by several individuals. Christopher Lozinski, BPG's founder, will be responsible for supporting, and extending the BPG BLOCKS Workflow engine. Dr. Bob Smith, one of the first customers for BPG BLOCKS, will be working for BPG as project manager. Jason Williams, one of BPG's early technology clients will be working for BPG as a Taligent developer. Jonathan Koyler, who first worked with BLOCKS in 1991 will be working as a Developer on the BLOCKS software.

These are all individuals who have demonstrated a long-term commitment to working in these technical fields. These people, will be working for BPG, which since 1988 has been steadily investing in the BPG BLOCKS manufacturing framework. As a group we have demonstrated a long-term commitment to this approach to software.

4.a.6. Cohn and Marks

Cohn and Marks is a Washington legal firm specializing in telecommunications and intellectual property law. Charles Oliver, a Cohn and Marks partner, became interested in electronic property during prior work as Attorney for Federal Communications Commission Common Carrier Bureau as Director, Legislative and Regulatory Policy. He previously worked for CBS Inc as Senior Vice President, National Association of Broadcasters; and as Senior Policy Adviser for the Office of the Assistant Secretary of Commerce for Communications and Information.

Bill Booher was instrumental in developing the ATP program plan in his prior work with the Department of Commerce.

4.a.7. Suite Software

Suite Software is a privately held company with offices in Anaheim, California and Dallas, Texas. The company was founded in 1988 and provides products and software services for VAX/VMS, UNIX-based environments, Windows, OS2 and NT.

Suites Product, DOME, is an environment and tool kit for creating distributed applications which can be used from any point within the network, regardless of the hardware platform or operating environment on which those applications are running. With DOME, application resources (which are often called objects) can be shared by other applications on any platform within the network. This is possible by way of transparent network communication, provided by DOME Communication Services.

4.a.8. Coalition for Electronic Markets (CEM)

CEM was chartered in 1993 as a Virginia Corporation in order to establish a corporate vehicle for housing initiatives of this nature. Its founder, owner and sole employee is Dr. Brad Cox.

Dr. Cox has been active in the object-oriented software community since its inception. He wrote the book, Object-oriented Programming, An Evolutionary Approach, that is credited with launching today's industry-wide enthusiasm for object-oriented technologies. He cofounded the Stepstone Corporation a decade ago, where he originated the Objective-C^reg. programming language and Software-IC^reg. libraries.

The objective of Cox's work is bringing about a software industrial revolution in which software is produced, not by fabricating everything from first principles, but by assembling interchangeable (reusable) software components which are in turn supplied by lower-level echelons of producers. Apart from his role within CEM, Dr. Cox is a faculty member of the Program on Social and Organizational Learning at George Mason University, an interdisciplinary department that concentrates on ways of overcoming socioeconomic and technical obstacles to change, development and learning as organizations adapt to a fast-paced global information age economy.

4.a.9. TBD: Account Management Software

The time limits on this proposal did not allow us to initiate arrangements with a provider for a crucial component of any financially sensitive system, the user interface through which buyers and sellers register as new CEM customers, monitor account balances, and register new electronic goods as products.

Although we intend to keep the architecture open so that diverse interfaces by different companies can integrate with this system over time, we hope to add a coalition member whose financial account management package might serve as the interface between end-users and their financial account. Technically viable candidates for a user interface that could be extended to handle the (long term) end-user market are Quicken, Managing Your Money, and others. Suitable user interfaces for the (short term) corporate user market are not known at this time.

4.a.10. National Institute of Standards and Technology

We look forward to continued participation of the NIST groups that collaborated with CEM in its prior effort with the Department of Defense last summer to deploy an infrastructure of this nature.

4.b Resources Available

Although this coalition made great strides in forming this joint venture this summer, we did not succeed in establishing binding agreements as to staffing levels that members will make available to this project. This was not due to lack of interest nor ability, but lack of time. To some extent, it was also due to the concern of some members that Washington lacks the will to support truly paradigm shifting initiatives. The coalition members could rapidly resolve this shortcoming once substantial support for this initiative is demonstrated.

4.c Past Performance

Except for a similar proposal to the Department of Defense last summer with a subset of this membership. this proposal is the first venture of this coalition.

5. Broad Based Economic Benefits

The ATP solicitation specifically requested "paradigm shifting" proposals. Thomas Kuhn triggered the present interest in the term in The Structure of Scientific Revolutions. But since this term has been since diluted to the point that its meaning cannot be taken for granted, we include the following historical digression to show the magnitude of the challenge, and the opportunity, that CEM reads into this term.

5.a. Potential to impact US Growth & Productivity

The paradigm shift addressed in this proposal is providing the financial incentive that might support an information age economy. The closest historical parallel to an effort of this magnitude was America's prior transition from an craft-intensive to a standards-intensive manufacturing practice during the industrial revolution.

Two centuries ago, the U.S. was an unspecialized agrarian economy. Weapons (muskets) were built by generalist craftspersons in small shops where each gunsmith fabricated parts from first principles, rather than by assembling musket from interchangeable components of other craftsmen as is the rule in manufacturing today. The same is true of the software industry. Programmers fabricate the components of their applications from first principles instead of by assembling prefabricated, interchangeable components developed elsewhere.

Ely Whitney's pioneering efforts were never fully successful. Almost a generation after he tried and failed, Roswell Lee and John Hall succeeded at the Springfield and Harper's Ferry National Armories. The effort was sustained by financial support and unflagging encouragement from the U.S. Congress and Bureau of Ordnance.

Today this long period of failure is forgotten. The interchangeable component approach that arose from these efforts was the basis for America's rise from a colonial backwater to the dominant world power during the First and Second World Wars.

If information age industry is ever to achieve a breakthrough in software engineering productivity, it must begin a similar transition. We must leave behind the practice of highly-skilled craftspersons fabricating everything from first principles. We must adopt an assembly-based approach, like Armory Practice, where lower-skilled workers (even computer end-users) assemble large-granularity components from interchangeable off-the-shelf parts.

The obstacle is not technical feasibility. Innovations such as object-oriented programming and other comparable technologies have made interchangeable components possible already, at least within limited domains. The problem is the absence of a credible solution to the electronic ownership issues posed in this proposal, since without property rights for electronic goods, there will be no off-the-shelf components for others to reuse.

As the Age of Manufacturing declines, low-wage smokestack industry jobs move offshore. America's best prospect lies in the higher-paying work of an information age economy. The long standing dream of an information age workforce transporting electronic goods and services around the globe via a Global Information Infrastructure is technically feasible already. But for this politically fashionable dream to become financially plausible, we must deploy a credible solution to the electronic ownership issues posed in this proposal.

5..a1 Near Term

The problem with paradigm-shifting initiatives is that (a) they are intrinsically bound up in chicken versus egg dilemmas and (b) it is not possible to peer through the paradigm shift singularity to accurately predict how fundamental changes will unfold. This section addresses this by showing how the proposed solution would address widely recognized breakdowns within the software industry as that industry is constituted today. The next section will try to anticipate changes in the post-paradigm shift economy enabled by the proposed ownership mechanism.

Charlie Oliver of Cohn and Marks writes:

Computer software and other forms of computer-readable information are large and rapidly growing industries in which the United States has significant competitive advantages in the international trading arena. Rapid growth has masked the enervating effect that inadequate intellectual property protections have already had in that market. But the proliferation of network connections will exacerbate the problem if royalty collection mechanisms are not adapted to changing realities.

In 1992, U.S. vendors supplied 74 percent of the world packaged software market, achieving revenues of $47.6 billion[2]. U.S. vendors supplied more than two-thirds of the world market for computer-aided design, computer-aided manufacturing, and computer-aided engineering (CAD/CAM/CAE), an industry that generated revenues of $16.5 billion in 1993. Electronic information services grew 16 percent in 1993 to an estimated $13.6 billion, with U.S. providers deriving about 35 percent of their revenue from foreign sources. Data processing and network services increased 14 percent in 1993 to about $46.4 billion. The Bureau of Labor Statistics reported that the computer software industry employed nearly 435,000 people in 1993, and that information services -- defined as companies that generate, process, and distribute data -- employed more than one million persons.

Some of the same technological advances that have fueled the growth of these industries could impair their performance in future years. In 1993, 39 percent of personal computers worldwide were connected to networks, up from 33 percent in 1992. The advantages of networks for increased productivity are obvious, but networks make it easier to share information of all kinds - including computer software and computer-readable data. The potential for that kind of sharing presents vendors with less-than-perfect alternatives. If the vendor continues to price information or software on the assumption that one user will have access to it, the prospect that purchasers will share that material - with increasing ease, over networks -- represents a loss of potential revenue. If the vendor prices its products on the assumption that everyone connected to the purchaser will use it, the result could be to price the product out of the market.

Site licenses may be a viable alternative during the brief transitional period when users are connected only to local area networks (LANs), but LANs are rapidly being interconnected with each other. The worldwide market for internetworking devices grew 38 percent in 1993 to almost $2 billion. Advanced Network Services, Inc., the company that operates the NSFNet backbone, says that traffic on the Internet has been growing at 20 percent per month for the past three years, compounding monthly. More than 10,000 information networks - and perhaps about two million individual computers - are already connected to the Internet. With all of those users potentially capable of sharing proprietary information with each other, site licensing has become increasingly untenable as a solution to anything.

Behind the robust growth of the software and information services industry, there are disturbing signs of less-than-perfect health. Though the Internet is widely extolled as a limitless source of information, much of the shareware offered for free on it is about as useful as its price. Even when altruistic developers offer clever or creative software for free, they can rarely afford to provide the documentation and on-line user support than can be essential to effective use. Software offered for remuneration is increasingly sold through, and owned by, companies that can afford to support large marketing organizations. In 1993, Business Week surveyed the 19 largest software and services firms and found that two of them, EDS and Microsoft, accounted for 40 percent of total sales and 54 percent of total profits. Competing vendors are rapidly acquiring each other in an effort to achieve similar marketing advantages.

Software suppliers to the personal computer market increasingly offer suite pricing, in which word processing, spreadsheet, and database applications are sold as discounted packages. In the short run, consumers may benefit from the bulk discounts being offered for software packages, but the reality is that most are purchasing capabilities that they will never need. The typical word processing or spreadsheet program today is replete with exotic and seldom-used subroutines, included to ensure that no competitor can claim even a theoretical advantage.

At the same time, on-line information services are priced on the assumption that most users operate in a corporate environment where the material will be shared with many others. The result is that sole proprietors and small business people are priced out of the market. For software and other information services, the marketplace today is like a restaurant where food is available only as part of a five-course meal, with drinks included. That kind of establishment has its place, but in this case, it's the only one in town.

In the long run, consumers are better off with compensation systems that allow them to pay only for what they use, and industry benefits as well. Esoteric and seldom-used software routines may be quite valuable to some users, and those users should be willing to pay accordingly. If simple, commonly used capabilities are offered at reduced prices, they can reach larger markets. And if information is priced on the basis of how many people are actually using it, it can provide value to small businesses as well as giant corporations.

Phil Casini of Dallas Semiconductor writes:

As distributed LAN technology continues to grow in popularity, the complex licensing issues that have been introduced, as a result of software being distributed across LANs, is proportionally increasing. In many cases, the effect of downsizing is causing many companies to spend large sums of money to cover license fees from their vendor. This is a cash drain on corporations that often hurts their bottom line.

In recent years, floating and site licenses have helped ease some of the financial burden. Many software vendors are moving towards offering floating license schemes. Recent announcements by IBM, Novell, and Microsoft, in support of server based license management utilities (for their respective network operating systems) supports this trend. Microsoft in particular, has proposed a License Sever Application Programming Interface (LSAPI) standard.

As the floating license becomes more popular, executives can begin to measure the productivity of their software purchases and their return on investment. However, the floating license in its current form limits executives to only gross measurements. There is not enough data captured about the specific use of the license to gain any more insight into the software's use. And while the floating license does ease some of the financial burden, up front investments are still steep.

Software users are now beginning to ask for temporary floating licenses, or metering. This is a logical extension of the floating license, because metering can be implemented as a pay per use system. This could further reduce (or eliminate) the up front costs associated with initial software purchases, allow executives to further measure the utility gains of the software, and could create more business opportunities for the software vendor.

From an accounting perspective, metering represents a more predictable and even payment plan, which corporations can better budget for and manage. In addition, response to peak demands can be better managed. Many hours are often spent today (particularly in engineering and manufacturing departments) creating ad hoc licenses for temporary software usage for extra personnel assigned to a project.

This form of license is a problem for software vendors, because the customer wants to use the software for a limited time, and cannot justify the payment of the entire license fee. In essence, the customer wants a license with a fee prorated for the length of the peak demand. As companies move to some form of a virtual corporation (or at minimum downsize their staffs) and jockey their resources to maximize their business goals, these situations will become more frequent. Time based metering solves this problem. Metering also allows executives to better measure their investments because specific information is recorded, perhaps by department and user. Licenses that expire with finer granularity than months or years also help executives control expenditures in response to down turns in their business.

Software vendors also stand to gain from metering. In some cases, high value software may suddenly become affordable for a small or midsize company with a limited cash flow. This scenario represents a sales increase for the software vendor, because that small company would not have made the purchase otherwise.

Metering also helps software vendors realize a more steady cash flow into their company, rather than the "burst" sales from a new product or release, with marginal sales revenue in between. Whether time or usage based, metered software represents a method of commerce that both customers and vendors can better manage. The model better fits the way corporations operate, budget, and manage a business.

The CEM proposal is particularly strong because it borrows concepts from traditional commerce practices, while introducing new technologies which allows the commercialization of the concept. These new technologies lower the cost of implementation to affordable levels, which has been one of the roadblocks to implementing metering in the past.

The software industry is dominated today by small number of companies with sales concentrated in a few large horizontal applications. Word processors and spreadsheets dominate the end user application business. The industry is largely composed of general purpose tools selling as commodity items where price is the most critical factor. This type of market rarely rewards innovation and the providers of niche market solutions find it difficult to compete. Component-based software has the potential to provide a "level paying field" by permitting smaller Independent Software Vendors (ISVs) to effectively compete. Companies, such as Taligent, Inc. and others have demonstrated that independent components coming from the ISV community can be composed into integrated solutions that uniquely address customers' needs.

Revenue collection for fine grained components remains a major obstacle. There is little incentive for an ISV to develop components and distribute them unless there is an adequate compensation mechanism. Revenue collection on today's large horizontal applications is a recognized challenge with piracy being the major form of fraud in the industry. A robust component-based software industry is dependent on a revenue collection mechanism being developed to address small, fine grained components.

Today's mechanisms for revenue collection as exemplified by Licensing Servers, such as Gradient Technologies, Inc's iFOR/LS, are designed for large application products. Authorization to use a software product is separated from distribution. A key to use the product is provided to a user in exchange for financial compensation to the owner. Today's licensing servers are wholly inadequate to the task of licensing the use of numerous, independent fine grained components. In addition, to foster rapid growth of a component-based software industry, new terms and conditions will be required including usage or rental charges, in addition to outright purchase.

5.a.2 Long Term

No one expects electronic goods to replace tangible goods altogether. However within the last generation, we've all experienced the sense in which the global economy is transitioning from an Age of Manufacturing to an Age of Information. For example, this author was raised on a dairy farm in South Carolina, in a community where at least 80% made their living by building tangible agricultural goods for a living. This author now lives in a Washington suburb where at least 80% make their living from various forms of white collar work. A substantial fraction of these produce electronic goods in the sense that term has been used throughout this proposal, even in the absence of a solution to the ownership issues raised in this proposal. Within a single generation, the dominant fraction of the American workforce has left the farm (and recently, the factory). The workforce is increasingly engaged in intangible white collar work.

During this generation, we've seen the computer and telecommunication industries continue the trend of manufacturing since the industrial revolution. They've supplied computer technology and telecommunications bandwidth with phenomenally improving price/performance ratios. A similar trend has not transpired within white collar work, particularly the white collar workers we call programmers. During the same period that hardware manufacturing became famous for exponentially growing achievements, the software industry became famous for the software crisis.

With only rare exceptions, white collar work is bought and sold as "services" not "products". For example, except for the shareware `industry', most programmers market their talent as services for a salary, as distinct from building electronic goods to be sold as products. Most white collar work is done under the same basis, as service providers instead of as product providers. With an ownership mechanism such as the one proposed here, an indefinite proportion of white collar work might be provided in either manner. For example, programmers could choose between selling services to software companies for a salary and building reusable software components to sell as products. Some white collar workers, such as order-takers, would have the option of taking orders for companies on salary or working at home building "ValidatedSalesOrder" objects for buyers of such electronic goods.

The taxpayers seem to share President Clinton's enthusiasm for a Nationwide Information Infrastructure. However existing information infrastructures like America Online, Compuserve and Internet teach a cautionary tale. They warn that high quality content will not be assured by connectivity, regardless of the bandwidth. A water supply does require investing in the plumbing. But plumbing is no assurance of a water supply if the water bill doesn't get paid.

The dominant practice within all electronic infrastructure is that all revenue flows to the infrastructure provider. Apart from marginal enterprises such as shareware, none of this revenue flows to the subscribers who supply the content these infrastructures to carry. To see the implications, one only needs to contrast the commercial vigor of any neighborhood bookstore with the low-quality chitchat on the newsgroups and mailing lists of the electronic infrastructures of today. As the old saying goes, "You get what you pay for". What subscribers get for an internet subscription is access to infrastructure. But the content of this infrastructure is similar to a trash disposal dump, not the farmers market that this proposal would enable.

The gigabit infrastructure President Clinton envisions is clearly capable of great achievements. But mobilizing the vast creative energy of the American populace will not occur in the absence of a robust incentive structure, capable of encouraging corporations and ordinary citizens to invest their labor in providing high-quality information products like software. Although most of this proposal has emphasized large enterprises like the software and publishing industries, these industries are today small compared to what could be accomplished by mobilizing the creative energies of entire populations.

5.b. Timeliness WINDOW...not too early nor too late

Systems for collecting revenue for software are not new. The industry has considerable experience with revenue collection systems. These traditional approaches are called pay-to-acquire because they employed technology to inhibit acquisition of software until fees have been paid. First generation efforts did this with copy protection technologies such as dongles and ways of making distribution disks non-copyable. But these were emphatically rejected by the market on the grounds that they penalized honest users. Recently, CD disks have provided ways to distribute software freely that can be enabled with a telephone call and a credit card number. As distributed computing becomes increasingly common, the industry is now exploring network-based license server technologies in response to the revenue threat that distributed computing could pose. Two of CEM's members plan to announce license server systems within the next year. Microsoft plans to bundle a license server with the next release of Windows NT.

It is a common misconception that CEM's approach addresses this same problem. These approaches are focused on the needs of the software industry as that industry is constituted today. CEM's focus is longer term, redefining the software industry as the industrial revolution redefined manufacturing. In particular, CEM's goal is not just providing a infrastructure that would address the revenue collection needs of the large companies who today provide large-granularity components like computer application. CEM's goal is to also address the revenue collection needs of those who might build the reusable software components from which large applications might be assembled. The crucial difference is that existing approaches are unigranular, focusing on revenue collection for large granularity components like computer applications. CEM's approach is multigranular, supporting revenue collection for components of any granularity.

This leaves only two initiatives that we're aware of as similar[3]. The oldest (1983) is a Japanese project headed by Dr. Ryoichi Mori of the Japan Electronics Industry Development Association (JEIDA). He calls this approach superdistribution by analogy with superconductivity because it allows information to flow freely without resistance from copy protection or piracy. His approach relies on tamper-resistant silicon technology. Although JEIDA is a well-connected non-profit coalition of large Japanese companies, it is not clear whether this project has advanced beyond the research stage. The last article Mori published indicated that they had completed two iterations of silicon fabrication in 1990 and were starting on a third.

A similar initiative is underway in the U.S.. Peter Sprague, the chairman of National Semiconductor Corporation, founded Wave Systems in 1988 to bring a similar silicon-intensive solution to market. Wave is not CEM's competition but a prospective member. In meetings this spring, Peter Sprage expressed willingness to work with CEM in principle. However he was unable to participate this summer because he was occupied with an initial stock offering. On the other hand, it is not presently clear that Wave's approach is capable of supporting multigranular revenue collection. It was certainly not a capability of the sole demonstration application Sprague showed at this meeting, and small granularity components seems not to have been part of his thinking.

Quoting from a PC Week article[4] about Wave's approach, "Wave's business plan is to break even on the encryption software and the decryption/metering chip. The real profits, Sprague says, will come from royalties as customers access data. Wave wants to charge information providers roughly 35 cents on each sales dollar. Sprague says the company is willing to offer as much as half of its take to PC makers as an incentive to install the chip on their machines and on their computing devices." His offer of an incremental revenue stream should be increasingly enticing as hardware profit margins continue their decline.

Although CEM is not too late, it may be too early. The worldview of the computer industry is still dominated by the "Personal Computer" mentality. It is only beginning to transition to distributed interconnectivity. Internet, Prodigy, America Online and Compuserve are beginning to attract substantial attention, but most computers aren't even connected to modems, let alone LANs. The Nationwide Information Infrastructure is still more dream than reality.

The problem is that it is impossible to be sure if we're too early without actually trying. The only certainty is that if we don't do something to capture the promise of information age commerce, our global competition will do it for us. The competition the U.S. most needs to worry about are resource-poor countries, such as on the Pacific Rim, to whom electronic commerce should appeal because it requires no natural resources. Electronic goods are solely a product of the drive and energy of people.

These nations' situation today is similar to the situation the U.S. was in when it invested in radical initiatives that ultimately displaced Britain as the dominant world power. Established economic powers must ask themselves, "What if there is a silver bullet for the ownership issues of electronic commerce. And what if the competition builds it first?"

5.c. Why ATP support is necessary

Though each participating company independently has plans to improve the revenue collection process for software, these efforts are not currently "inplan" as other more pressing opportunities have usurped the existing funding. In addition, more iterative and independent efforts are the norm in addressing this kind of an opportunity, due in part to funding constraints in commercializing software. ATP funding provides an opportunity to bring development efforts forward in time and to design a more inclusive and robust solution by forming a cross industry partnership.

The completed and commercialized revenue collection mechanism for component-based software will foster rapid growth of components as the developers will perceive the opportunity to be fairly compensated for their efforts. System integrators and ISVs will combine these components into unique solution packages that can be used in whole or parts can be replaced by the user with their own preferred component tools. This replacement and combination opportunity will release the strangle hold on the software industry and will encourage the rapid growth of a components industry.

5.d. Benefits related to costs

The preceding discussion discusses the cost versus benefits of paradigm shifting initiatives.

6. Plans for Commercialization

The technical components of this proposal should not be allowed to obscure the fact that CEM's fundamental business is not technical but financial. CEM's business is like that of credit card operations such as MasterCard or Visa[5], not technology companies such as Apple or Microsoft. CEM's business is handling other people's money, conveying it from buyers to sellers of goods. Its incentive for providing this service is the fraction of each transaction it debits as its fee. From this it pays operating expenses and distributes as dividends and/or share value to its investors.

CEM is a registered Virginia corporation, chartered to house the work of deploying and operating this venture as a going commercial concern. Its initial investors are the companies participating in this proposal, who receive shares in proportion to their investment. However, although the value of these shares could become significant in time, this is not our member's primary reason for investing in CEM. Our members are established in other lines of business. Their primary early incentive for participating in this proposal is the value they expect participation to bring to established product lines.

Further details about the corporate structure of this enterprise, such as formulas to protect the ground floor interests of early investors while encouraging latecomers to join, have not yet been arranged. Similar arrangements are to be made regarding such questions as under what circumstances intellectual property developed during this project resides with CEM or the member institutions.

6.a. Elements of the Commercialization Plan

Although we anticipate considerable overlap and cross-fertilization between phases, the commercialization can be thought of as proceeding in three stages.

• Although this proposal has emphasized invocation over acquisition as a basis for commerce in electronic goods, we don't think that invocation will replace acquisition entirely, neither in the short run nor, in fact, ever. Therefore we plan to begin operations with licensing server technologies now being used by CEM's member companies. This would provide a marketable service to today's software industry, provide a testbed and market for technologies being developed for the invocation-based approach, establish a commercial presence in the revenue collection market, and establish the synergistic relationship between renting and owning that is the norm in other industries.

  This first stage will service the needs of the software industry as that industry defines itself today. Since today's software industry views the revenue collection issue in acquisition-based terms, mentally equating revenue protection with copy protection, this phase will concentrate on deployment of license server technologies. The target market is large companies selling large-grained components (computer applications) to large companies (as distinct from individual consumers). This stage involves initiatives that coalition members are planning to deploy independently of this proposal. It will commence this fall and continue in parallel with subsequent stages.

• The second stage deploys an invocation-based revenue collection infrastructure as a synergistic alternative to acquisition-based license servers within the same market segment targeted in Stage 1. CEM's competitive advantage is the ability to allow software producers to offer terms and conditions that are precisely tuned to the needs of their customers. They do this by providing CEM with terms/condition algorithms that convert invocation counts to financial amounts due. These algorithms can support any terms/conditions, ranging from pay-to-own, to pay-if-you-like-it, to radical incremental charging options where the fee paid is precisely proportional to the software's measured utility to the end-user. Since the target market of this stage will be large companies (in which inhibitions to fraud are strong), the metering infrastructure will be weighted towards software over hardware. The main hardware component will be a Smartcard(TM) or Button(TM) device which will be used authenticating the individual who's actually incurring the charges (as distinct from computer serial numbers, which identify computers not users).
• The culminating stage expands invocation-based charging from the trunks to the branches and leaves of the national economic tree. At each stage in this progression, from large organizations to smaller companies to ordinary individuals, increasing levels of technological sophistication will be needed to keep ahead of the growing likelihood of fraudulent behavior by both users and vendors.

CEM understands that, under ATP guidelines, inplan activities are not eligible for ATP funding. Therefore this commercialization will emphasize commercialization plans for Stage 2 and subsequent activities, while linking them to the commercial presence that will result from Stage 1 market development activities.

6.a.1. Customer Targets

The target market for initial deployment is large organizations who buy software applications for use by their employees. The organization that represents these interests is the Society of Information Management, the professional organization for more than 2,700 senior information technology executives around the world. Its 1992 member survey identified software licensing as a critical information technology issue. The following quotations are from SIM's Software Licensing Position Statement of April 1993 in which this society summed up the view of software consumers within large corporations of means for resolving their difficulties with software licensing.

The ideal software vendor and customer partnership of the future will be based on an understanding of mutual issues, trust and the need to balance revenue stream generation for vendors with cost control for customers. As organizations move toward multivendor, multiplatform, open systems environments, the pricing models of today must give way to more flexible and equitable licensing models and allow for more regional and global focus on the part of customers.

The evolution toward rapid application development technology further adds to licensing fees. Licensing models have not kept pace with this evolution. The evolution of desktop software towards the use of object-oriented technologies will transform the "Application-Centric" environment to one which is "Document-Centric." This will further add to pressures for licensing which addresses software integration trends.

Vendors must have the ability to monitor customer usage. Fair and equitable licensing models should ensure vendors' revenue streams are providing cost advantages to customers. [Emphasis added]

The emphasized text shows that SIM's approach is consistent with that of our proposal. SIM's president, Steve Coppell, responded to CEM's request for an endorsement for the proposal CEM submitted to ARPA/TRP last summer as follows:

"The Society for Information Management (SIM) has addressed changes and issues regarding software technology and recognizes the software monitoring issue as one of the obstacles preventing this technology from moving forward. SIM is closely watching with interest the proposal put forth by the Coalition for Electronic Markets."

6.a.2. Access to Customers

During initial deployment phases, CEM must reach two classes of customers, large end-users for computer applications, and the vendors for those applications. It must also provide a way for component providers and users to acquire the metering infrastructure in a painless fashion, so that the buy decision is not complicated by requiring them to spend money just to sign up.

Taligent plans to install the invocation-based metering infrastructure in the operating system they'll provide to their major investors (Apple, IBM and Hewlett Packard) and other equipment manufacturers. Hardware-intensive instantiations (Stage 3) will deploy in an analogous fashion, by having the metering components integrated into computer hardware as it comes from the factory, with incentives to the hardware manufactures provided by offering a share of the incremental revenue.

With the metering infrastructure in place, end users will be contacted via traditional marketing channels or via professional organizations such as SIM. They'll be invited to compare their software costs via the two options that will then be at their disposal; paying for software as they do now and paying for software via the metered infrastructure.

This will of course be preceded by a comparable effort to "stock the shelves" with metered software. Software vendors will also be offered a similar choice, continuing as today versus metered distribution. Most vendors will chose to do both since the cost of developing a metered version of each product is small; a matter of two subroutine calls. They'll choose to invest this small cost because the savings in distribution are so large, using electronic distribution instead of costly current distribution channels, avoidance of some advertising via software that can advertise itself, lowered decision to buy threshold via "try before you buy" terms and conditions, and so forth.

6.a3. Customer decision to buy knowledge

Although it would be premature to rule anything out at this early date, CEM does not believe it should obtain its operating revenue as customers are acquired. Our deployment plans are based on the assumption that customers (buyers and sellers alike) will acquire accounts and thus access to the metering infrastructure without fees. The "decision to buy" the infrastructure should be as close to zero-threshold as we can possibly make it. CEM expects its revenue to come when its infrastructure is used, not when it is acquired.

Therefore the "decision to buy" we're primarily concerned with occurs when a potential customer encounters a component. This might be in a free download area on a network, in a diskette a vendor mails to its customer list to advertise a new product or release, or (as the information economy starts to develop), in higher-level "components" that work just like physical software stores, but which are run by entrepreneurs who choose worthwhile components, organize them attractively, and sell them electronically for a markup collected via CEM's infrastructure.

A requirement on every vendor by the terms and conditions CEM imposes on every infrastructure users, is that the price of every transaction be made known to the customer before the charge is committed. This might be done by physical labels on diskettes or via splash screen advertisements within the code. Vendors are responsible for insuring that the price charged (by the terms and conditions algorithms) not exceed the price advertised. Such overcharging can be detected by any customer (by examining their account balance), and if reported to and confirmed by CEM, will be dealt with severely as fraud.

Therefore the decision to buy is easy for the end user. Traditional transaction costs, including the need to drive to a software store or dial a catalog store, are reduced to essentially zero since the software is already here. Assuming market forces have operated as we expect, the invocation-based price should be markedly lower than the consumer's acquisition-based alternative, so there is no barrier to trying it. And of course, for software distributed under pay-to-use terms and conditions, there is none of the uncertainty about whether the software will be useful that we all experience with pay to own purchases.

6.a4. Customer support requirements

We anticipate that the customer support requirements will be comparable to those of other financial operations such as credit cards. That is, huge in principle, but reduce in practice by CEM's ability to leverage networked technology for major operational cost savings. After all, point of sale operations are entirely electronic. Even downstream consumer activities, such as reconciling the monthly bill, is electronic. This account interface (see Figure 2) is accomplished via software functionality such as Quicken (or a comparable package oriented to the account management practices of large companies during Stage 2). For these reasons, the operational costs of CEM's system should be much lower than with credit cards. Similar savings should be possible on the supply side. The administrative tasks of registering a new product also occurs electronically, via capabilities to be supported within the same account interface package.

The primary operational staff requirement arises from the need to handle cases where the mainline electronic system breaks down, handling violations of terms and conditions (elsewhere discussed as `fraud') being the one likely to lead to significant staff requirements.

6.a.5. Strategic adaptation to market dynamics

CEM's approach can be thought as techno-centric front end appendages within users PCs and back end (financial account management) appendages within existing financial institutions. These appendages are coupled by the core of CEM's business plan, the Clearing House.

Although the appendages will receive the most attention during development stages of this deployment, this core is the heart of CEM's business. The appendages are expendable and can be replaced as market dynamics dictate by competing enterprises. The clearing house is not expendable for it is where CEM draws its revenue. In return CEM supplies the service of integrating small transactions reported by the front office appendages into transactions the back office financial institutions can process cost effectively.

6.b. Technology Diffusion & Protection of Intellectual Property

CEM's proposal is a means of protecting electronic property (a subset of intellectual property) by providing a mechanism through which revenue can be acquired by the owner as electronic goods are used. Preliminary investigations into protecting this approach under patent law[6] concluded that U.S. Patent Law does not provide patent protection to "ways of doing business". Furthermore, there is a strong consensus within CEM's member institutions (articulated most strongly by David Sharp of Taligent and John Slitz of the Object Management Group), that market infrastructures should be open. By this we mean that implementations can be proprietary, but interfaces must be documented and that competing implementations welcomed so long as they comply to documented interfaces.

CEM's intention at this time is to protect this system as credit card companies protect theirs, not by owning the fundamental concept but by establishing a market presence sufficient to withstand the pressures of an open competitive market.

7. Proposer's Level of Commitment & Org Structures

CEM made significant progress this summer obtaining firm commitments from its members, as will be evidenced by the commitment letters attached to this proposal and in the sections that follows. However we failed to arrange the binding commitments we'd hoped to provide in this solicitation. The primary obstacle was the time required for advocates within companies to understand the magnitude of a paradigm shifting initiative, to communicate this understanding to decision makers in their companies, and to develop quantitative data capable of competing effectively within short-range priority lists.

We believe that this is not an unusual problem but is characteristic of all joint ventures in the formative stage. We also believe that this uncertainty is especially characteristic of paradigm shifting initiatives specified by the ATP guidelines. This should not be construed as evidence of any disinterest in deploying this approach commercially.

7.a. Level of Commitment

Since some of the corporate commitments to this proposal are non binding at this moment, CEM decided to provide verbatim statements from coalition members in the following sections, followed by brief mention of areas that are tentative. We believe that binding financial commitments can be achieved quickly once ATP demonstrates its own interest in our approach.

7.a.1. Taligent

David Sharp supplied the following letter of commitment for Taligent (also see attached letter):

Taligent believe this partnership and the revenue collection proposal can make a significant contribution to the realization of a component-based software industry and marketplace. We look forward to a successful partnership with our fellow consortium members and the Department of Commerce on this project.

Taligent, Inc. is pleased to be an active participant in the Revenue Collection Infrastructure for Component-based Software consortium. We are excited about the prospect of our advanced, object-oriented system becoming a platform for the development , distribution and use of component software including a revenue collection mechanism that will encourage and stimulate the deployment of new solutions and tools. We believe a revenue collection and distribution infrastructure will materially enhance America's electronic marketplace not only for computer software but also for all manner of electronically distributable products such as movies and books. Taligent believes this will improve our nation's competitiveness and are proud to be part of this effort.

Taligent, formed to provide the operating environment for the next generation of desktop computing, has considerable expertise and resources to accomplish this task. Our preference is to develop the proposed revenue collection and distribution infrastructure in concert with other partners as this provides the opportunity for the broadest short term market acceptance. However, Taligent is prepared to fund the development of a similar but more narrowly defined revenue collection and distribution mechanism on its own to address explicit requirements coming from our developer and user communities. Working in concert with others and with the assistance of ATP funding will broaden and enrich the scope of such efforts and ensure a more ubiquitous deployment.

Taligent believe this partnership and the revenue collection proposal can make a significant contribution to the realization of a component-based software industry and marketplace. We look forward to a successful partnership with our fellow consortium members and the Department of Commerce on this project.

Several unresolved issues are implicit in this letter. From Taligent's viewpoint, the ideal way to proceed is as outlined in this proposal, with a coalition of large companies capable of deploying a broadly based system. Also implicit is a fall-back plan that we describe as Taligent-centric. If we fail to achieve broad buy-in from other companies, Taligent is prepared to invest sufficient money to deploy an analogous (but radically scaled back) plan that solely addresses the needs of Taligent investor companies. This fall back plan would involve other present members depending on financial commitments resulting from continuing negotiations within CEM.

The ATP guidelines draw a very fine line by demanding, on the one hand, paradigm shifting initiatives, that on the other hand, are not something companies already plan to do. But successful companies abide by Carnegie's ascerbic observation, "Pioneering don't pay". CEM has struggled to provide a plan that abides by ATP guide lines, but the uncertainty about where this line really lies has plagued us from the beginning.

7.a.2. Object Mgmt Group (OMG)

John Slitz provided the following letter of commitment:

Object Management Group (OMG) plans to participate in the CEM group for the following reasons:

1) OMG has launched an electronic distribution system -Information Brokerage (IB)- for information and software about object technology, component software and distributed computing. This program must confront the issue of licensing and revenue collection that looms over its activities. The CEM approach to this area can greatly assist the growth of the market for electronically distributed information and software by relieving concerns, by sellers and buyers, about software piracy, fair pricing for use, and accounting for sales and use of electronically distributed products.

2) IB will work with the other members of the group to define the requirements for the CEM system as they would be most useful to the IB. IB will also act as an early, open, world-wide test bed for the new technology.

3) IB will participate in a technical capacity in the development of the system at such a time as deemed necessary and helpful by the other group members to make the CEM licensing mechanism available on the IB system.

4) OMG believes that the commercial success of this project will rely as much on the packaging and marketing of the technology as it will on the technical abilities of the developers to meet the many technical challenges in the development of the actual technical deliverables. To that end OMG sees the participation of the IB as a natural place to accumulate requirements and test ideas for various packaging alternatives that arise from the technical development. OMG as the largest worldwide consortium of software developers and producers in the world, would seek to keep its members informed and solicit participation by interested parties to broaden the support for the CEM solution.

The OMG has been in existence since April 1989 and currently represents over 430 members world wide. In responding to the requests of its members and the technology producing and consuming public for faster access to information and software, it has formed the Information Brokerage. IB will commence operations in January 1995, with a world wide launch of its services. The focus of the IB is the technical users and producers of object oriented software. This target market is, for the most part, expert in the use of computers and telecommunications technologies on which the IB is built. Little training or education will be necessary once a user has logged on to the system. Terms, conditions, and pricing of products and services will all be left to the judgment of the provider of the service or product.

By introducing the "usage based" pricing paradigm implicit in the CEM proposal IB management believes that a degree of order and standardization to what promises to be a dynamic market condition. If the CEM mechanism leads to substantial adoption of this approach it will undoubtedly stimulate substantial growth in this segment of the software market.

The same issue of inplan versus paradigm shifting is also implicit in this letter. OMG has an inplan initiative underway to deploy license server technologies as soon as Fall 94. At the same time they recognize that acquisition-based license servers cannot address the long term requirements of commerce in component-based software.

7.a.3. Citibank

Daniel Schutzer provided the following text to include in this proposal

At a minimum Citibank would like to participate in the ATP in the following way: To observe and study the social acceptance and dynamics of selling software (and any commodity) on a usage basis rather than on a per unit basis. We would also like to explore the various pricing issues and strategies associated with these two competing approaches. If it looks like a sound business strategy we will enter and help develop a commercialization plan and perform the function of the bank. As a side interest, we intend to review the security and encryption system design aspects and apply our expertise in fraud control and detection to help validate the security and integrity of the offering.

Dan Schutzer and Frank Bamberger are within the Research and Development arm of Citicorp under Colin Crook, Executive Vice President of Research and Development. Although they are personally committed to the vision expressed in this proposal, they must persuade organizations outside the R&D branch to make financially or operationally binding commitment. Both have worked, and are continuing to work, to achieve these commitments, but they are not in hand at this moment.

7.a.4. Dallas Semiconductor (DS)

Phil Casini, Director of Marketing, proposes DSC's role in this project as follows:

Dallas Semiconductor will participate in the ATP proposal in the following roles:

* Technology provider of the metering hardware. Buttons are proposed as the tamperproof billings time source and possibly the transaction repository.

* Offer consulting engineering services to help those partners who will develop the metering and billing software rapidly integrate Button interfaces into the appropriate software modules and make the most effective use of the Button devices.

* Limited role in developing the proposal(s) from a marketing perspective. This means I can make my self available to help develop portions of the proposal that would benefit from marketing expertise and to participate in planning meetings from a project management perspective.

Michael Bolan, a DSC founder and a member of its board of directors, represented Dallas's interests within CEM during its proposal to Department of Defense last summer. During that solicitation, which was on a much reduced scale, Dallas committed to an in kind investment of $480,000 to be matched by ARPA funds. Although no comparable commitment letter is available for this solicitation, CEM views this as evidence that comparable investments can be forthcoming once planning uncertainties are resolved.

7.a.5. Berkeley Productivity Group (BPG)

Christopher Lozinski, BPG's founder and president, provided the following letter of commitment:

BPG is pleased to be an active participant in the Revenue Collection Infrastructure for Component-based Software consortium. BPG is able and willing to commit 115 hours per week at $100.00 per hour for the duration of the project. This represents four different individuals work, three part-time, and one full-time. We are willing to provide up to this level of investment, providing that the project work load requires it.

Since 1988 BPG has been trying to establish component software as the basis for the next generation of manufacturing systems. Finally this year, the market has accepted the technologies required to make this happen, and BPG's business is taking off.

Meterware is an important new application area for the BPG BLOCKS class library. Our customers are already very interested in being able to track their manufacturing costs in an object-oriented fashion. Tight tracking, and careful analysis will allow them to reduce their manufacturing costs. Applying the same principles to controlling software costs is a natural extension of our business.

Lozinski adds, in response to Casini's contribution to Section 5.a.1 of this proposal:

Thank you. I now understand much better what we can do for the financial managers of the world. They need to allocate software costs to the things that they sell. If they sell projects, then they need to track the resources used in the project, including the software in order to allocate costs appropriately. If they sell products, then they need to allocate the software usage to the product being produced. BLOCKS can help do this. BPG BLOCKS models the relationship between production resources, and the products produced. In this application it can be used to relate meterware tracking costs to the work being produced.

I now understand why Bob Smith has been pushing for BPG BLOCKS in this proposal. It is not just so that BPG BLOCKS can be superdistributed and charged as meterware. It is because BPG BLOCKS models the organizations workflow, and can be used to track software usage. BPG BLOCKS is the meter engine that allows the financial manager to allocate software costs to particular projects.

During CEM's summer 1993 response to a similar Department of Defense solicitation, BPG committed to an inkind investment of $272.035, to be matched by federal funds. Although no comparable commitment was offered for this solicitation, CEM views this as evidence that comparable investments can be forthcoming once planning uncertainties are resolved.

7.a.6. Cohn and Marks

Charles Oliver proposes Cohn and Marks' contribution as follows:

Draft a joint venture agreement that would provide incentives for participating companies to contribute money, personnel, and intellectual property to the development of software, communications systems, and organizational arrangements necessary to support a billing and collection system for use of software. Assist in writing project overview explaining commercial purposes of project and benefit to the public. (Oliver was a principal author of open systems rules requiring telephone companies to provide interconnection with enhanced service providers, and he prodded the FCC into launching its expanded interconnection rule makings to provide interconnection with competitive basic service operators.) Translate proposals of engineering types into language that will be understood and accepted by business and legal types. Assist the project team in laying the groundwork for marketplace standards that will be broadly accepted without running afoul of antitrust laws. Assist the project team in selling its systems to major government and commercial users sufficient to create momentum toward broad acceptance of the system developed. Advise the project team on opportunities and problems that have arisen over the years in the implementation of comparable systems, including ASCAP and BMI.

Apart from Oliver and Booher's personal contribution to the coalition building and proposal writing effort, which were substantial, their firm's involvement in the coalition is likely to be as a subcontractor as distinct from as an investor.

7.a.7. Suite Software

Ken Fleming, Suite's founder and president, provided the following commitment letter for this proposal:

SuiteSoftware is pleased to be an active participant in the Revenue Collection Infrastructure for Component-based Software consortium. Since 1988 SuiteSoftware has been trying to establish component software as the mainstay for distributed systems. Our commitment is on the same level as Brad Cox's namely a lifelong commitment to making this happen. Some customers level of commitment is equally strong. Encompass is using an object oriented, component based architecture for their Global Trade Management System. Chaparral Steel is a one of our manufacturing customers that is using a component based architecture. Many of our customers share our vision, and will continue to support this concept. These are all people with a technical vision of where the market is going, making business decisions according to their principles.

As this moment, CEM has received no information regarding the possible level of Suite's financial commitment.

7.a.8. Coalition for Electronic Markets (CEM)

This proposal was written under the assumption that CEM is to be the corporate home for this venture. Other options are possible that would change the organizational details of this proposal, but not its substance. For example such an option would be to house a joint venture within one of the larger members of this coalition.

In either event, we anticipate that development and operation of the Clearing House component of this proposal would be managed by Brad Cox. Two options are available for achieving the financial wherewithal for implementing this responsibility, which might be used in parallel.

One is to fund the Clearing House as Cox founded his prior company, Stepstone. This was to initiate development with personal capital, launch the business with a private stock offering and leverage growing revenues with venture capital.

The second option is to involve Dr. Cox's contacts within George Mason and other universities to mobilize the research talents of graduate students and faculty. This would, of course, be done on a subcontractor basis as specified by ATP guidelines.

7.a.9. Other Potential Members

Many other companies have expressed interest in and support for the coalition's objectives. Its present members were selected who managed to respond within the limited time available for preparing this proposal. We are especially grateful for the encouragement and support of the following individuals and institutions and look forward to their broader participation when time constraints are less pressing.

• Steve Koppel; Society for Information Management
• Bob Terry; NASA AdaNet Repository Project
• Jim King; NCR Corporation
• John McGehee; Texas Instruments
• Paul Mattson; IBM AS400 Division
• Peter Sprague; Wave Systems
• Dave Roland; Roland Projects, Inc.

CEM owes a special debt of gratitude to Paul Kozemchak of the ARPA Policy Planning Office. Without the proactive encouragement and enthusiasm that he so generously provided, we'd have never undertaken this project.

7.b. Type of Structure

The diverse technical, social, legal and business challenges of this project requires close cooperation among business, academia, and government. The Coalition for Electronic Markets (CEM) is a corporation chartered as the vehicle for addressing the ownership and control issues that will emerge as this project unfolds, for raising follow-on capital, and for deploying this system as a commercially viable business.

Ownership of CEM will be via stock to be issued in proportion to investment, with appropriate weighting (to be determined) favoring ground floor investments while not excluding follow on capital raising efforts.

The present membership was selected according to skills that span the major architectural components of this system, from silicon fabrication to software to telecommunications to financial operations. Another major component of the selection process was also that some companies with strong interest could not respond within the time frame of this proposal. The roles of each member company is discussed in [[section]]4.a.

7.c. Participation by Small Business

Three of the coalition members are small businesses, Berkeley Productivity Products, Suite Software, and the Coalition for Electronic Markets.

7.d. Sub Contractors and Other Business Alliances

The participation of Cohn and Marks and George Mason University would be as subcontractors, the latter as required by ATP guidelines.

7.e. Management Plan

Overall project management will be by Brad Cox of the Coalition for Electronic Markets. Responsibility for silicon fabrication is with Dallas Semiconductor, the workstation-resident infrastructure software with Taligent, central clearing house with CEM, and financial account management integration and services with Citibank.

The level of commitment within the coalition is varied at present, as is typical for paradigm shifting initiatives. A common source of uncertainty is that we've found it hard to distinguish between the strong commitment of the individuals who have worked on this proposal and their managers from that of their companies as a whole.

8. Budget

Due to time pressure and the inexperience of some of CEM's member companies at responding to proposals of this nature, figures are available only for the larger components of this system.

        Source of Support           1995  1996          1997          Total            
        A. ATP                $2,445,000  $2,718,750    $1,600,500    $6,758,750       
        B. Participants       $2,445,000  $2,718,750    $1,600,500    $6,758,750       
        C. State/Local                $0  $0            $0            $0               
        D. Other Federal              $0  $0            $0            $0               
        E. Other                      $0  $0            $0            $0               
        F. Total              $4,890,000  $5,437,500    $3,201,000    $13,517,500      

8.a. Breakdown by task for each year of the project

Task                        1995  1996          1997          Total            
Infrastructu            $707,500  $787,000      $447,250      $1,941,750       
re                                                                             
Financial               $616,000  $695,500      $435,250      $1,746,750       
Integ                                                                          
Silicon Fab             $303,500  $338,750      $135,500      $773,250         
Clearing                $818,000  $897,500      $582,500      $2,297,000       
House                                                                          
Total                 $2,445,000  $2,718,750    $1,600,500    $6,758,750       

8.b. Breakdown by cost category for each year of the project

Cost breakdown figures did not arrive from the members by the deadline for this proposal

8.c. Totals by Participant by year

        Partner/Contractor              1995  1996          1997          Total         
        1. Taligent                 $707,500  $787,000      $447,250      $1,941,750    
        2. Citibank                 $616,000  $695,500      $435,250      $1,746,750    
        3. Dallas Semi              $303,500  $338,750      $135,500      $773,250      
        4. CEM                      $818,000  $897,500      $582,500      $2,297,000    
        5. Total                  $2,445,000  $2,718,750    $1,600,500    $6,758,750    

9. Bibliography

Many of these articles are available electronically at /mon/CEM

1. Adams, Sam S. 1992. "Software Reuse and the Enterprise," unpublished manuscript, Knowledge Systems Corporation.
2. Barnes, Bruce H. & Bollinger, Terry B. 1991. "Making Reuse Cost Effective" IEEE Software, January.
3. Biggerstaff, Ted J. & Perlis A.J. 1989. Software Reusability V.1 & 2, Reading, MA: Addison-Wesley.
4. Boehm, B. W. 1982. Software Engineering Economics Englewood Cliffs, NJ: Prentice-Hall.
5. Brooks, Frederick P., Jr. 1974. The Mythical Man-month: Essays on Software Engineering Reading, MA: Addison-Wesley.
6. Brooks, Frederick P., Jr. 1987. No Silver Bullet: Essence vs. Accidents of Software Engineering Computer 10-19, April.
7. Cox, Brad J. 1986. Object-Oriented Programming: An Evolutionary Approach Addison-Wesley {Objective-C++ is a superset of C++ in the sense that Objective-C is a superset of C. The compiler was developed by NeXT Computer Corporation as an enhancement to the Free Software Foundation's GNU C++ compiler, and is distributed by both NeXT and FSF.}
8. Cox, Brad. 1990. Planning the Software Industrial Revolution IEEE Software 25-33, November.
9. Cox, Brad J. 1992. What if there is a silver bullet and the competition builds it first? Editorial; June 1992; Journal of Object-oriented Programming. Subsequently republished in the October 1992 issue of Dr. Dobb's Journal.
10. Cox, Brad J. 1993. Market Processes as a New Foundation for Software Engineering; Under review for the 1993 ACM SIGSOFT Symposium on the Foundations of Software Engineering, chaired by Barry Boehm.
11. Cox, Brad J. (forthcoming book) What, if Anything, is an Object?, Addison-Wesley.
12. Cusumano, Michael A. 1991. Japan's Software Factories: A Challenge to U.S. Management New York: Oxford University Press.
13. Eidelman, J. A. & Shepard, C.R. 1986. Living among Pirates: Practical Strategies to Protect Computer Software Michigan Bar Journal March.
14. Felton, Eric. 1990. "The Hard Road to Software Copyrights" Insight Feb. 19.
15. Furubotn, E.G. & Pejovich, S. (eds.) 1974. The Economics of Property Rights Cambridge, Mass.: Ballinger.
16. Hayek, F.A. 1945. "The Use of Knowledge in Society," in Hayek, Individualism and Economic Order Chicago: University of Chicago Press, 1948.
17. Herron, Keith D. & Witt, Joanne T. 1991. "Microcomputer Manager's Association White Paper on Network Software Licensing," Infoworld Oct. 14.
18. Kuhn, Thomas S. 1960. The Structure of Scientific Revolutions University of Chicago Press.
19. Lavoie, D., Baetjer, H., & Tulloh, W. 1991a. "Coping with Complexity: OOPS and the Economist's Critique of Central Planning," Hotline on Object-Oriented Technology 3: 1, (Nov) pp 6-8.
20. Lavoie, Don, Baetjer, Howard & Tulloh, William. 1991b. "Increased Productivity Through Reuse: An Economist's Perspective" Proceedings of the Third Annual Workshop on Reuse Software Productivity Consortium, Herndon, VA.
21. Lavoie, D., Baetjer, H., Tulloh, W. & Langlois, R. 1993. Component Software: A Market Perspective on the Coming Revolution in Software Development, Special Research Report, Patricia Seybold Group, Boston, MA.
22. League for Programming Freedom. 1990. "Software Patents: Is this the Future of Programming" Dr. Dobb's Journal November.
23. Lorenz, Mark. 1991. "Real-world Reuse" Journal of Object-Oriented Programming Nov./Dec.
24. MacKaay, Ejan. 1990. "Economic Incentives in Markets for Information and Innovation" Harvard Journal of Law and Public Policy Summer.
25. Meyer, Bertrand. 1990. "The New Culture of Software Development" Journal of Object-Oriented Programming (Nov./Dec.)
26. Miller, Mark S. 1989. "Charge-per-Use Software" unpublished Manuscript.
27. Mori, Ryoichi & Kawahara, Masaji. 1990. "Superdistribution: The Concept and The Architecture" The Transactions of the IEICE 73:7, July, Special Issue on Cryptography and Information Security.
28. Nagle, Thomas T. 1987. The Strategy and Tactics of Pricing: A Guide to Profitable Decision Making New York: Prentice-Hall.
29. Pool, Ithiel de Sola. 1990. Technologies without Boundaries: On Communication in a Global Age Cambridge, Mass.: Harvard University Press.
30. Pratrap, Sesha. 1992. "Achieving Component-based Programming Today" Object Magazine Jan/Feb.
31. Robertson, Paul L., & Richard N. Langlois. 1992. "Modularity, Innovation, and the Firm: the Case of Audio Components," in Mark Perlman, ed., Entrepreneurship, Technological Innovation, and Economic Growth: International Perspectives Ann Arbor: University of Michigan Press.
32. Society of Information Management. 1993. "Software Licensing Position Statement," April 1993
33. Taylor, David A. 1992. "Whatever Happened to the Object Components Industry?" Object Magazine Jan/Feb.