Next: The Libre Services model Up: Libre Services: A non-proprietary Previous: Introduction Contents
The Internet has created an enormous new offshoot of the software industry: the Internet services industry2. By Internet services, or Subscriber Services, we mean any service that is provided to a user via the Internet. Some examples are:
And many other types of service. The above list is far from complete, nor is our simple categorization of services in any way definitive. Subscriber Services is a new industry, not yet 20 years old, and still undergoing a process of disorganized self-definition. Thus far evolution of the industry has been driven by a multitude of ad hoc commercial initiatives, and it remains in a state of cheerful chaos.
Nevertheless, the scope of what we are describing is extremely large. In effect, it is the entire global Internet itself.
Though still undergoing chaotic evolution, the Subscriber Services industry is well established. In 2006, generalized Internet services are provided by several large providers such as AOL, MSN and Yahoo. These major providers deliver a broad range of services to their subscribers, including most of the examples given above. Google is also aggressively moving into the generalized services arena, rapidly augmenting its core search service with a variety of additional services.
In addition to the major providers of generalized services, there is a wide variety of providers of specialized services, such as classified advertising (Craigslist), auctions (Ebay), airfare and vacation booking (Expedia), job listings (Monster.com), dating (Match.com), car trading (AutoTrader.com), and numerous others. We can expect that over time the large general service providers will provide many of these specialized services too, so that the industry will consolidate into a small number of dominant providers.
In addition to the variety of new services that the Internet has enabled, a fundamental change is occurring in the way traditional software applications are being provided to the user. Before the appearance of the Internet services industry, software applications were either run locally, on the user's own PC, or perhaps on a remote server on a local area network.
But now traditional software applications are migrating towards a service-based implementation. Many user applications that hitherto have been implemented as local stand-alone applications, are now being implemented as Internet services, in which the user is provided with the same or similar functionality via the Internet.
This represents a fundamental shift in the focus of the software industry. The focus is moving away from software as a product, and towards functionality as a service. This trend is just now becoming widely recognized within the industry, and there remains confusion about its exact nature and implications. This general confusion is reflected in the multiplicity of terms used to refer to this trend, such as ``software as a service,'' ``information technology as a service,'' and ``transformation of software into services.''
All these developments--both the evolution of the Internet services industry, and the migration of traditional software into a service-based implementation--are taking place almost entirely in a proprietary context. Even Google, despite any pretensions it may have to the moral high ground, is based on proprietary software, heavily defended by patents, copyright and trade secrecy.
In addition, the Free Software Foundation and other organizations exist to provide leadership and advocacy, and to serve as rallying points for participation. As a result of all this, the free software movement exists as a viable alternative to the proprietary model.
However, a corresponding set of non-proprietary formalizations does not exist within the services domain. Today, virtually all Internet services are provided under the traditional proprietary software model, and as yet the free software movement has no formal presence in this domain. A set of defining principles has not been established, nor is there any leadership or rallying point for participation in the services domain, corresponding to the leadership of the Free Software Foundation in the software domain.
As a result of this there does not exist a non-proprietary alternative to the proprietary services model.
Meanwhile the proprietary services industry continues to grow rapidly.
It also continues to undergo a process of consolidation. In the services arena, as in the general software arena, there are strong forces of convergence towards a small number of dominant providers, and eventually a monopoly. In the more mature software industry this has resulted in the Microsoft Windows monopoly, now with no proprietary competition at all.
The services industry is already undergoing a similar process of convergence, and this may continue until there is a monopoly in this arena too. Microsoft is a particular concern in this regard. With its dominant position in both user environments (in the form of Windows), and services (in the form of MSN.com), Microsoft can create a level of integration between its own proprietary user environment and its own proprietary service that cannot be matched by any of the other major players. Based on its operating system monopoly it can exercise dominance of the services arena in a way that the other providers cannot, and can eventually achieve a monopoly position there too.
This trend towards a commercial oligopoly or monopoly presents some societal concerns.
The Internet is a global public resource, and as such requires representation and advocacy for the public interest. Such representation includes maintaining technical standards and protocols, protecting civil liberties, and preventing unfair business practices.
In the technical arena of standards and protocols, public representation has historically been provided by standards organizations within the Internet industry, such as the Internet Engineering Task Force (IETF). But service providers are under no obligation to use public protocols, and are free to deliver services using their own proprietary protocols. And if a commercial provider is large enough, in terms of number of subscribers, it becomes a de facto authority within the industry, able to establish its own protocols as an industry standard.
The history of Instant Messaging is a good example. Clearly the right answer for Instant Messaging is a set of non-proprietary public protocols, allowing global interoperability and open participation by all service providers.
But this is not what happened. Even though IRC (Internet Relay Chat) already existed as an open protocol for Instant Messaging, the major providers disregarded this entirely. Instead, each defined its own set of proprietary protocols, in an attempt to take complete ownership of Instant Messaging functionality and deny this functionality to competing providers. The result was that each provider created its own island of subscribers, unable to communicate with the subscribers on the competing islands.
Only when the resulting dysfunctionality became unacceptable to their subscribers did the major providers address the issue. But not by adopting an open protocol--rather, by implementing interoperability gateways amongst one another. The major providers now have cross connectivity for Instant Messaging, but the underlying protocols remain proprietary, and smaller providers are marginalized and at a competitive disadvantage. In effect the major providers have created a cartel to take complete ownership of Instant Messaging.
The presence of the free software movement in Instant Messaging remains limited to a small island of IRC users, who are able to communicate with the proprietary subscribers via interoperability gateways provided by the proprietary service providers. The major providers thus continue to control the Instant Messaging arena, with the free software community provided for as a courtesy.
Thus in the technical arena such public representation as exists is ineffective. The large commercial providers are able to establish their own standards and protocols, and engineering standards organizations such as the IETF have become largely irrelevant.
In the area of civil liberties there is no formal public representation at all, and protection of civil liberties remains entirely in the hands of the commercial service providers.
If the commercial providers could be relied upon to act in the public interest, this would be of no great concern. But such is not the case. The commercial providers are under no obligation to protect the public welfare. Their sole mandate is to pursue their own commercial interests, and these may be highly detrimental to the broader interests of society.
Two recent events demonstrate this very well. In the first event, Microsoft recently shut down the Internet journal of a Chinese dissident, who had been expressing political views that his government found objectionable. In the second event, Google recently agreed to censor its search results in China, expunging web content that the government considers objectionable. Google will base its censorship decisions on guidance provided by Chinese government officials. So much for the moral high ground.
Both of these actions were taken at the request of the Chinese government, and both companies complied because this was in their commercial interests: they did so in return for access to the Chinese market. This was a quid pro quo arrangement, in which a political favor was exchanged for a commercial one.
Despite the simplistic justifications offered by their respective public relations departments, the fact is that Microsoft is silencing freedom of speech, and Google is degrading freedom of information. These are clear trespasses against basic civil liberties.
These are stark illustrations of how commercial interests can be highly injurious to the human condition. Any doubt we may have about the hazards of entrusting Internet governance to the commercial providers is surely dispelled by these examples.
In the general software arena, the free software movement and GNU/Linux play an essential role in providing a non-proprietary alternative to the Windows monopoly. It is imperative that a similar non-proprietary alternative be established for the services industry. Without such an alternative, Internet governance will remain largely in the hands of commercial interests.
All the basic principles of the free software movement carry over into the services domain. In particular, Internet services are amenable to similar cumulative and collaborative development mechanisms to software. And as in the case of software, these mechanisms depend critically on the appropriate freedom of action.
But in order for these mechanisms to be replicated in the services arena, the proper formalization is required. This formalization must include: a coherent model for non-proprietary services, a definition of the concept and principles, creation of industry-wide awareness, a framework for collaborative development, leadership, and a rallying point for participation. In short, an equivalent movement to the free software movement is required in the Internet services arena.
