Netart, Internet portal, Directory, Forum, Cybernetics, Disintermediation, Collaboration, Art, Public sphere, Information architecture, Hierarchy, Open source, Interactivity, Search engine, Content filtering.

• Disintermediate the art selection-evaluation-categorization process in the domain of netart
• Produce an extensible open source forum/portal
• Deconstruct prevailing portal architectures
• Limit the influence of marketing over information content
• Distribute control of the portal

In the framework of the Internet portal, the role of the public is conditioned by software architecture and interface design. The most visited and influential Web directory, for example, makes little use of the potential contributions of the public. It has an opaque content evaluation process -- apparently conducted by temporary workers with dubious insight into the appropriateness of their decisions. This has proven to be a rigid, conservative model because it is less expensive to ignore the evolution of various fields of knowledge. The emergence of the Open Directory Project as a major source of categorized information about Web content is suggestive of the potential of collaborative filtering (www.dmoz.org). Unfortunately, this model for database maintenance, which relies on volunteer editors, has not generated particularly high quality information about quickly evolving fields like online art. Despite its rapid early growth, many aspects of the ODP are developing at a glacial pace. Like its commercial counterpart, Yahoo!, much of the information it offers is out of date. Looking at the ODP and Yahoo! sections relating to netart, it is clear that they have not grown to reflect the current complexity of the field. 

Compared with existing alternatives, a more dialogic and revision-friendly portal is imaginable -- one that allows users much more preferential control over the presentation of the database's contents. There are significant strides to be made in both interface design and in the technical integration of mail-, Web-, database-, and messaging-protocols.  This project, however, addresses more than just problems of technique and interface design. 

While a few existing netart portals have developed some encouraging characteristics, none has pursued the democratic footing appropriate to an influential volunteer project. Given that artists and critics currently contribute without remuneration to provide the majority of the content of existing netart portals, moving towards an "open" alternative seems to be a justifiable paradigm shift.  Using the ODP as one of its models, the proposed portal intends to involve many people in suggesting and transforming online content.  The intent is to establish a collaborative online institution whose function is clear and compelling so that volunteers will emerge to manage and develop it.

Moreover, unlike the ODP, it is possible to make the portal's code as available as its 'core' databases, so that the overall system will be truly public and transparent in nature. In accordance with trends in open-source software development, the project's software will be most viable if it charts a course for general utility. Its component-organized structure is conceived with a view towards applicability to other fields of knowledge. This should inspire others to intervene in its codes, to help them weather the inevitable changes in Internet protocols and browser functionality. With this in mind, the proposed portal for netart can be thought of as the first implementation of a new collaborative filtering and forum system.

Similar to the governmental principle dividing church and state, Altport is constituted to resist abuses of power; which for netart could involve promotion, or allegiance to particular artists or media-formats. Like the Open Directory Project, Altport does not try to function as an archival institution. There is no art collection and no licensing of artwork. 

A portal should not have to dabble in archiving. Effective archiving is complex and resource intensive. Moreover, in this world so saturated with advertising and marketing, the spectre of self-interested promotion hangs over a portal that involves itself in the acquisition and 'preservation' of online art. Most prototypic netart portals straddle the line between information about "their" resources and information about other websites. While such hybridization may not always be sinister, Altport aims to avoid this functional ambiguity. The Internet itself forms the relevant boundaries to its world-wide "collection." 

Although it will not archive netart, Altport will include documentary text, iconic images, hyperlinks and a variety of related data contributed by artists, critics, users, and the general public.  

In general, Altport is concerned with diverse representations of databased information. With a few exceptions, present portal alternatives insinuate a market-organized order that, however arbitrary, nevertheless begins to seem reflexive. That "Autos" has crept to the top position in Netscape's Netcenter directory, supplanting "Arts," is suggestive of this problem. Unlike the ODP, Altport will not offer a one-size-fits-all interface.

There will not be one branching structure that greets all visitors. Rather, people will articulate the stance they wish to take vis-a-vis the portal, selecting their specific interests: which topics are of primary interest, whose opinions matter, etc. The interface is conceived as a flexible, personalized representation of the portal's "core" database. Although this idea is not without precedent, it has barely begun to manifest itself in existing portals.

For several years now, particularly since the demise of Adäweb, Rhizome has become a standard of comparison for netart portals. Rhizome is one of the few portals that has resisted conventional hierarchical representational practices: it's not an option to navigate through top-level topics to find subordinate topics in Rhizome's Artbase. Instead one uses keyword searches. But this is not a functional panacea, as any number of biased search engines demonstrate. Representational biases are displaced to the less obvious realm of algorithms and database design. The categories chosen by artists for their art "objects" are factored into the results of each search. How they are factored depends on indexing algorithms. Ultimately, for any given keyword, everyone receives the same results in the same order.

Though meant to be a feature, the absence of a visible hierarchical navigation interface can be considered a flaw. Rhizome seems to be inhibited by the perception that hierarchies are un-rhizomatic. The problem is that imagining spatial arrangements for ideas has been a proven and effective mnemonic device at least since orators developed the technique in ancient Greece. The desire to resist hierarchy, and to liberate diverse lines of flight, is thus set against the practical utility of the portal itself as an aid to memory. If hierarchical categorization poses a problem in portals, it is because too often particular hierarchies are imposed and they are coercive. Confronting this problem does not require that visualization be abandoned. Altport makes navigation hierarchies configurable. Users affect their own interfaces and fluidly share their alterations. Participating in the process of revision, people are likely to recognize that hierarchies are subjectively produced. 

Not merely pragmatic, giving people more control over the way they structure their information "navigation,", mnemonic associations, categories, and filtration strategies will offer, through its difference from a range of existing portals and online interfaces, a new experiential vantage point from which to understand familiar representational patterns and biases.

One fundamental problem in designing a more user-directed portal is the identification of meaningful and intuitive choices that will effect the contents of the portal. In a field like netart, with myriad announcements and public discussion threads, forms of information filtering are of obvious practical importance. Altport addresses the challenge of categorizing and documenting quickly changing and politically contentious material. When filtration requires nuanced decisions, these are most effectively made by humans. But who will make these decisions? If opinion formation in netart has been anti-democratic, this undoubtedly owes, in part, to the laboriousness of salvaging intelligent discourse from the rabble of online culture. More than merely abundant, elements of noise can be difficult to avoid.  For years subscribers to electronic newsletters have chosen between impossibly noisy unfiltered subscriptions, which include every message sent to a list-serve, and more manageable weekly summaries -- often compiled by the same person every week. There comes a point when this is not a job for one person. It's important to give people choices concerning who will filter new material and categorize content. Between noise and conventionally curated material, a variety of meaningful subscription "portfolios" can be articulated whose character can be determined largely by users.  

Altport does not simply give users new ways to shelter themselves from an overabundance of information. Altport's capacity to conform to the user's preferences is important, but equally important are the ways that people can be involved -- more or less casually -- in processes of comparison, evaluation, criticism, association, selection, and documentation. This is not an improbable Utopian dream that promises to founder on the rocks of anarchism. Many existing projects demonstrate the potential energy of public participation: book evaluations at Amazon, music affinity matching at Napster, member reliability ratings on Ebay, discussions at Net-time. It is evident that volumes of useful information can be generated spontaneously by interested participants. Still, finding a popular model for governing such participation is not easy. Currently the decision-making that governs the evolution of online portal architectures is conducted almost entirely outside the public sphere. This, together with the evident influence that portals exert, invites reflection on political models of governance.

Altport proposes a variety of checks and balances. Insofar as is possible, preference options will be used to give people meaningful choices about what types of information they wish to see. But this is not enough, because any popular forum will quickly be exploited for unintended uses. This is where code architectures become critical. Mechanisms must be available to limit the circulation of "spam" email to users, but there must still be a context for the interchange of ideas. The privileges of anonymity and freedom of speech must be balanced against the rights of real participants. The power of one person using the portal must be balanced against the expectations of others. The interface should not elevate jabbering fanatics to a tyrannical status.  With regard to balancing signal and noise, much can be learned from what has already been done.  Problems can be minimized by good code, and good code follows from clear ideas. Care must also be taken to lay the groundwork for resolving disputes among users.  Policies as well as code are necessary.  The possibility must be left open for "constitutional" reform of the system -- primarily through alternative codes. The development of an excellent portal is something that will take time and revision, and this is all the more reason to begin now. It is likely that policies, interface solutions, and algorithms can be found that offer a much greater role for public participation than is currently common.

Indeed, most of today's portals are fortified -- they are designed to deflect controversy and criticism. They impose a variety of constraints that warrant consideration. Even outside the obvious bastions of online commercialism, interactive conventions have drifted toward an embrace with a subject who is ever more passive and influenced by promotion. With commercial software development focused on spectacular graphics and e-commerce, effects are felt in the realm of art, where online interactivity is framed by conventions that are ever more televisual. 

As a sector of software development, online art and culture (at the periphery of entertainment) are quite small. Already commercial design conventions have become the standard poles by which even open source software authors orient themselves.  Yet it is outside the mainstream currents of software in the dynamics of open source development that the arts are most likely to find willing and able allies. The move toward collaboration is a bulwark against obscurity for content that is not heavily advertised or intrinsically viral. While mercenary zeal and cynicism have fueled the evolution of portals and portal paradigms in recent years, there is an opportunity to develop an alternative portal that synthesizes the strengths of existing portals, and builds on the foundation of existing open source codes. 

Like bad laws, software can be rewritten. The politics of participation, access, ownership, and authority must be understood as chosen rather than as technical accidents. Online architectures that favor institutional and professional authority can make the difficulties of public participation inevitable rather than problematic. Online information ecologies need not replicate the models that have been commonplace in museums and galleries, let alone on commercial television. There remains the potential to expose tired old habits of cultural legitimation rather than recreating them online. Despite the abundance of promotional and irrelevant messages, the sitting reserve of Internet surfers have much to contribute; their voices and actions should have a place in the classification, filtration, and critique of netart.