Leslie Walker's 1/3/02 Washington Post technology column describes how an international law firm is utilizing a custom online collaboration system to enable interaction among its lawyers and clients around legal transactions, regardless of location or time of day. Okay, you're probably saying, "so what?" at this point. After all, that's what online message boards, newsgroups, chat rooms, and instant messaging have been doing for years. The difference here, however, is that the software at work allows documents and files to be accessed locally from individual PC desktops while sitting on other machines and servers across a network. The company behind the technology actually implemented an earlier version of the same technology with a major legal publishing corporation, according to an October 2001 InternetWeek. The touted advantage of the system, in essence, is that previously developed paperwork and documentation can be called up directly by any user, without having to resort to e-mailing large cumbersome files in potentially incompatible formats, guessing which versions of documents are most relevant, or addressing security issues peculiar to a particular local- or wide-area network. Now, if you're thinking that this database-driven distributed approach to knowledge- and content-management sounds awfully similar to a certain peer-to-peer technology used to trade music files, which itself has been the focus of much legal activity during the past two years, (and coincidentally was developed at roughly the same time as the aforementioned system in 1999), the irony was not lost in either the InternetWeek piece, or an 11/5/01 MSNBC profile While the law firm system's architecture does not operate under a voluntary content submission model-- and does not involve copyright issues around the content, unlike Napster, arguably the most famous peer-to-peer tool, there are some basic features the two share:

• First, the system itself is not physically limited to the servers within a particular local-area network: it enables each individual machine tapped into the system to become a server unto itself, thereby allowing for direct point-to-point content access across multiple locations, through a common interface.
• Second, nearly every piece of information around each documents placed within the system is captured. So rather than simply hunting for a piece of information by filename or date, users can locate resources by function, type, or whatever classification is desired.
• Third, online archive directory listings, which point to specific file locations on other individual hard drives, can be generated around a particular document or set of documents, in order to provide a quick reference point for related material. Where Baker & McKenzie are extending the application of the archive, however, is to use it as a means for developing tutorials and practica for associates or new parties who need to get up to speed on a particular case. Rather than redoing documents from scratch, new persons unfamiliar with a particular case or set of legal issues can simply access draft language, background material, or templates to fit a particular task. The law firm's setup does differ from Napster, though, in one key area: this setup is expensive. The firm's base system runs a reported US$85,000 for 250 users (the firm supports 3,000 lawyers and legal staff worldwide); the customization runs about US$350 for each user.

Peer-to-peer is a distributed information-sharing model, predicated on the idea that more types and levels of information can be provided from more sources, eliminating concerns around one server being the sole repository of information. Individual machines can function as both information hosts and clients, opening up opportunities for multiple users to both collaborate on and interact around documents as needed. The advantage is that documents can be easily accessed without resorting to cumbersome e-mail or time-consuming downloads, and without worrying about security or performance issues around an individual server. Nearly every piece of information around each document placed within the total file-sharing network is captured, enabling participants to access specific files by a wider range of indices-- which in turn point to specific file locations on other individual hard drives. The law firm, in question, is also utilizing the archive as a means for developing tutorials for new parties involved with particular litigation, so that they might be able to access draft language, background materials, or templates to fit a particular task. It's interesting that, despite the price of the proprietary architecture behind it, the law firm's use of peer-to-peer is focused not so much on simple document sharing as training and support. If there were any immediate application for peer-to-peer in a nonprofit environment, it might be in this area of training as well. Though we couldn't immediately locate any cheap peer-to-peer tools that might easily accommodate this, consider some of the courseware tools utilized in online distance education frameworks. The American Distance Education Consortium and Canada's Centre for Curriculum, Transfer and Technology's List of Comparative Analysis Applications provide useful starting points for some of the more popular collaborative courseware tools (a number of which are free). Consider the basic components of courseware tools: some are focused on online discussions and forums, others are setup to provide tests and quizzes, some allow for presentations of documents (including multimedia) to be displayed, others provide the means for real-time discussion among participants. The more advanced will combine all or some of these functions. There are two interesting (and free) systems that might prove a useful starting point for P2P integration, one proprietary and one open-source. The proprietary system, WebMentor Author, allows for the development of training modules. The content can be crafted on the fly from existing snippets of HTML and graphics already sitting on your machine. The core content can then be rendered as a self-contained interactive online learning environment, incorporating multimedia, applets, script, and other Web technologies. One interesting aspect is the ability to present content in a dynamic manner, including tests and quizzes, based on individual responses or skill levels, as well as tracking/monitoring individual progress in real time. There are two for-fee add-on modules that enable the course content to support real-time chat, and publishing of individual courses across a large-scale network and distributable media. The second system is called ClassWeb. This is an open-source, database-driven, web-based system, developed by the UCLA Social Sciences Computing group to provide a means for administering all aspects of a course, including distribution of class materials, announcements, message board discussions, online quizzes, registration and enrollment rosters, and grade management. It has been in continuous use to support some 300 courses at UCLA since its official release in 1997. So here are at least two models and tools that, if fitted with P2P capacity, might offer a means for collaboration with a relatively low barrier for participation among various users across different platforms, regardless of computing capacity. But would anyone truly undertake peer-to-peer for nonprofits? For all of it's benefits, the peer-to-peer model underlying Napster had issues itself. If a machine storing certain files is not switched on, or there are software issues on that machine, the files it houses are inaccessible to others. If the tasks are left up to individual users for labeling and providing information as to the files and their contents, without some sort of standardization and routine maintenance, the search index becomes outdated if not unreliable, and requires end-users to often examine the actual files themselves to determine its usefulness. And depending upon the volume of files and the frequency with which they are accessed or updated, there may honestly be no advantage in distributing files versus placing them on one host server, other than end-user perception. That said, if there were a system for nonprofits to provide access to a wide range of information-- such as mailing lists, organizational contacts, sample files and templates--- that match specific needs and interests without having to rely on file size restrictions through e-mail, or security issues, there might be some merit to a peer-to-peer approach. By underlying the file sharing with an online database backbone, those files, furthermore, could be enhanced by online commentary and ratings from end-users, and statistics around their access and use, which could also serve as a means of improving the quality of files available to others. If you're interested in experimenting with, and possibly contributing thoughts towards the development of, open-source approaches to peer-to-peer technologies, arguably the best place to start is O'Reilly's Open P2P Directory. This is a growing compendium of organizations, companies, individual experiments and initiatives. This includes direct links and descriptions to efforts (including some familiar ones) involving file sharing, distributed computing and search engines, instant messaging, advanced collaboration, and more. So, is anyone building the Nonprofit Napster? Resources Cited 1/3/02 Washington Post, Leslie Walker NextPage October 2001 InternetWeek, 11/5/01 MSNBC Napster American Distance Education Consortium List of Comparative Analysis ApplicationsCentre for Curriculum, Transfer and Technology's WebMentor Author ClassWeb O'Reilly's Open P2P Directory