Prepared by

Do not copy, link or reference without contacting the authors for permission

Introduction

What are the key, important, cutting edge exemplars of online learning technology that exist in North America context at the beginning of the 21st century? One way to detail advances in online learning is to survey the woof and warp of online courses offered: K-12, post secondary, graduate and others. This approach would provide a snap shot of what is presently offered, but answers none of the questions regarding where the technologies are heading, and what is currently possible. This report chooses the path less trodden. Our goal is to identify the key educational technologies out there on the Internet, without regard to their present use. That is, many of the topics that we take up are educational in focus, but the examples are often not. The chief 'mad scientist' at zeroknowledge.com was shocked to learn that his company's product could even be considered for educational purposes, and admitted that it offered a whole new avenue for their work. The goal of this report is to capitalize on key technologies and approaches, and to examine their uses in and as educational environments. Many of the applications of these technologies are as novel for their creators as it is hoped they are for the readers of this report.

In this report we have grouped our ideas into seven main areas: Community, Narrative, Constructonism, Economic Models, Archiving, Modes of Interaction, and Design. These are determined based on grouping of the key technologies, not according to a pre-existing theory of organizing online environments. These areas represent locations with the greatest potential to impact on learning environments in the near future. The importance of each area will be outlined in turn.

Our notion of community informed by the ideas of Ivan Illich, Elinore Ostrom, Seymour Papert and Howard Rhiengold. Illich, in his seminal work Deschooling Society, identifies community as self-teaching entity. In Governing the Commons: The Evolution of Institutions for Collective Action, Elinor Ostrom highlights self organizing and self governing communities centered around common pool resources, taking up issues of membership, participation and monitoring of community involvement. Papert extends Piaget's notions of constructivism into the technology friendly constructionism through the notion that we learn by building things and sharing them with others. Finally Rheingold asserts that online communities that can generate a sense of emotional response always pull participants towards real life interactions; you're not a community until you've attended your first funeral. The key issues are: reciprocity, sharing, action research, performing, telling stories. And the context is that these actions occur beyond overt or explicit economic arrangements.

The following description of GNU and open source comes from Nolan (2001).

There is an entire software community, multiple communities actually, that have existed on the Internet, along with many others, since its inception, predating the commercialization of the space in the early 90s. This community is presently represented in movements such as the Open Source, GNU and Perl initiatives. GNU and Open Source describe different philosophies of how to share information online.

Open Source and the GNU Project are two organizations influenced by specific individuals; GNU by Richard Stallman, and Open Source by Eric Raymond. They are now the champions of Free Software Foundation and the Open Source Initiative respectively . In general terms, they both want to promote software that is free, freely available, and open to the Hacker community. These projects both support the traditional notion of sharing resources among members of a community (Ostrom, 1990).

The two main camps, the Free Software Foundation and the Open Source Initiative, seem to be matters of personal taste to me, but they do have practical differences. According to Raymond, his "Open Source Initiative is a marketing program for free software. It’s a pitch for `free software’ on solid pragmatic grounds rather than ideological tub-thumping. The winning substance has not changed, the losing attitude and symbolism have" . Raymond’s project has the upper hand, I think, in that it is not tied to an explicit political or philosophical agenda, it is moving toward a standardization of what the idea of sharing software and code means in a manner that is palatable to industry, and it has a cooler name. Accordingly, it also has more commercial support especially by Hacker publishers like O’Reilly who publish many books on free and open source software http://opensource.oreilly.com/ . The Free Software Foundation, on the other hand, is older and broader based in terms of how it can be used. I can, and have put my own invention V.A.S.E. under a Free Software Foundation GPL license. I did not need to ask for permission. I only needed to download the text that would identify this license in my code, and abide by the license myself, in relation to this code.

These initiatives, and software specific initiatives focussed on languages, programs and specific environments such as Perl (http://theoryx5.uwinnipeg.ca/CPAN/perl/pod/perlfaq1/Who_supports_Perl_Who_develops_it_Why_is_it_free_.html), Apache (http://www.apache.org/LICENSE.txt), MOO (http://lingua.utdallas.edu/encore/), all share in the notion that strength and diversity comes from sharing what has been learned and done with others. The code is not the commodity, though the knowledge and wisdom sometime are commodified. The code is free. But if you don’t understand it at all, you can pay for the service and support.

The Internet grew out of the DARPA and the programming communities at a number of educational institutes. And of these early communities the programmer community of the various open source initiatives have grown. As well, various online communities have grown predicated, not on similar programmerly interests, but on communal interests. The earliest of these are the WELL and TWICS (Kroll, 1992) in which individuals of diverse backgrounds and interests coalesced into solid long term communities. Both WELL (http://www.well.org) and TWICS (http://www.twics.or.jp) represent online subscription services. The key issues for education are that these communities are ones of sharing of personal experience and knowledge that are intimate and long term. Reinghold writes of the WELL, showing how these informal learning environments supplemented or replaced the institutional support networks. The revelation is that groups of like minded individuals provide a more satisfying learning and communal experience than what can be found in formal learning environments. And that people are willing to pay for these experiences. It is interesting to note, that in some contexts, such as the WELL, that as the technology changed, from BBS to web based forums, the community moved with the technology, indicating some independence of the community from the technologies used.

Egroups.com, recently bought out by Yahoo.com adds to the notions of community fostered by the WELL. They provide tools for electronic community via email discussion groups that anyone can create on their own, and populate with friends, or via themes that allow individuals to control and govern their own community involvement. They also provide various tools like questionnaires, file uploads, and chat functions, but the key issue is that the control over the variety and types of communities are not governed by egroups.com, but by members as a free service, without tying members to a specific fee-for-service like AOL or various Internet Service Providers (ISPs)

Examples of theme or subject-specific communities that could be found through egroups, or other technologies are the Lucy Maud Montgomery discussion list (http://www.upei.ca) or the Serbian community (Nolan 1998, Nolan, In Press) where individuals interact around a particular literary, social or cultural theme.

Ability Online (http://www.ablelink.org/public/default.htm) is a community that encourages participation from children who have been marginalized by society due to illness (acute and chronic). It has been able to show that participation in their community has helped develop face-to-face social skills. Ability Online is supported by The Hospital for Sick Children (http://www.sickkids.on.ca), one of the largest pediatric, academic hospitals in the world. Another example is FAST (Food Allergy Survivors Together, http://www.foodallergyz.com/) which is able to provide support for food allergy sufferers not only through shared experiences but also by giving new allergy sufferers a forum to exchange recipes and product alerts.

Notions of performativity are important as locations for learning in cross/inter/trans cultural groups where it is how the individuals position themselves that form the location for community. The most obvious are writer's groups such as Inkspot (http://www.inkspot.com) and NetAuthor (http://www.netauthor.org). These two groups, one based in Toronto, and the other in Texas, are major players in the field. Inkspot has a monthly readership of 75000, and NetAuthor was listed 13 of the top 50 places to get published on the Internet. Both are sites run by writers, for writers, and on the topic of writing. The idea that writers don’t need to take a course in writing, but find a community of writers and work with each other building their craft and sharing skills, forums for publication, and secrets overshadows any sort of environment that a formal institutional program of writing could offer. And unlike professional writing societies, these groups recognize their mentorship role in helping new readers, and provide an opportunity for the established writer to provide a service to new writers starting out.

One of the key ideas that grew out of the early Internet is the notion of hacking, a philosophy that is underrepresented in learning initiatives. It speaks to constructionist learning, mentioned elsewhere, but partakes of the desire to perform a ‘return to service’ (Drake, 1993) and share of knowledge, skills, experience, and curriculum. Unlike the popular media presentation of hackers as the evil threat to the Internet, hackers are what brought the Internet into existence. Anyone who hacks at code for the explorative act of learning how things work, and shares their learning with a community, and are known for this act as a hacker IS a hacker.

Hackers, the people who break into computers, are technically Crackers; a term coined by the Hacker community in the 80s . Hackers are people who hack code to improve it. Hackers create things, while Crackers destroy things.

There is nothing malicious about hacking. It is a quest for knowledge of how things work, and it is a sense of community. This informal learning environment is what keeps unix sys-admins up to date on the latest bug fixes and defenses against viruses, worms, security holes and crackers. Though there are diverse forums that have sprung up on the Internet for communicating on hacker issues, the most important, and oldest forum is Usenet. Usenet, like the Internet itself, has no formal home or location. As with all the early features available on the internet until the advent of the WWW, such as email, telnet, ftp, chat, talk, finger, Usenet has been part of the communications foundation on which the present internet is based since the late 1970s (Krol,1992; Moore, 1994). Usenet is a news service of discussion topics that at last check included over 20000 different topics and discussions. They are organized into various hierarchies that include science, technology computers, social topics, sociology, broadcast news and anything people want to discuss in an open, mostly unmoderated, environment. Again the key issue is that of public, ad hoc, uncontrolled sharing of ideas, information, facts, fiction, fraud and everything in between.

To access a Usenet group, one only needs recourse to an ISP, and a software client. Web browsers can usually function as Usenet news readers. Almost all but the most conservative ISPs provide full access to Usenet news groups. Up until recently, the web portal to Usenet was via DejaNews (http://www.dejanews.com), but this service has been only partially available since DejaNews was bought out by the search engine giant Google (http://www.google.com)

As mentioned previously the Cracker community is vastly different than the Hacker community. Cracking is about breaking into computers, stealing passwords or code, and boasting about the exploit for power and prestige among the cracker community. If you want to know how to break into a computer, or bring down a major online corporation with the well known "ping of death" you find about it at cracker sites like http://www.warez.com for cracking; and the other location for sharing information are Internet Relay Chat (IRC) where there are thousands of hidden chat forums where people, mostly young males, ask about how to do things, tools for cracking various sites, and boasting about exploits. This is also a place where security and police agencies monitor activity, and learn for themselves how to thwart activities. Cracker communities are ad hoc informal learning environments where the content and communicators are unknown quantities, but where the judicious searcher for knowledge can find out information on a variety of nefarious topics.

The key importance of noting the Cracker community is two-fold. First of all, it is the place where many young males learn about technology. It is a place where they will study, learn and share knowledge even when they are not participating in formal classroom or online environments. The entire issue of interest and motivation come into play when students spend hours engaged in online learning of something that is meaningful to them, and will not spend seconds with imposed content. But the other point is that the Cracker community describes a learning and knowledge hierarchy that recognizes knowledge as power and privilege, and a constructionist learning environment (Foucault, 1979, 1991; Illich, 1970; Pappert, 1991).

The most important learning activity of the Internet is found in the telling and sharing of stories in various narrative forums. This capacity of the Internet is underutilized by learning institutions who are primarily focussed on controlled curriculum and easily accessible learning experiences.

The strongest form of personal narrative is the web page itself, and an astonishing number of free sites exist for hosting of web pages. But the most novel tool for the sharing of stories has emerged in the blogger movement (see http://www.larkfarm.com/weblog_madness.htm http://www.rebeccablood.net/essays/weblog_history.html by Rebecca Blood for more information). Blogging is the art of web journalling, or web logging. Web logging is contracted into the notion of Blogging. Blogs are narrative utilities that allow individuals to manage and present daily journals or logs. Blogger.com is the most famous, hosting over 60000 blogs. Blogger.com provides a suite of utilities that allow any one with a web page to maintain their own journals with the minimum of fuss and technical skill. After subscribing to Blogger.com, an individual, or group, can create one or more blogs. These blogs are accessed daily and track postings, manage archives and layouts, and then post the daily output to the individual/group’s web page. While not providing web hosting, per se, sites like Blogger.com facilitate a novel form of communication that harkens back to traditional forms of journalling found in many languages and cultures.

One level up from the uneven space of the blog are forms of epublishing where submissions are edited, usually with a weekly feature article about a variety of topics that are edited, and clearly semi-professional modes of discourse with a consistency that brings week after week. An example of this type of site is A List Apart (http://www.alistapart.com). Another example is Fray.com which accepts submitted stories, edits them for online publication and then allows readers a forum to respond to a specific question at the end of the article in an unedited format. There are even forums for epublishing, zines such as netauthor.org which are about epublishing, and literally thousands of online magazines for publishing everything and anything.

These forms of information and experience based narratives partake of the oral tradition of learning. The claim to authority is based on long term internal consistency of the work. While some blogs are personal accounts only, others are more theme-specific. Qube and Librarian.net provide a personal account of what's happening in their respective industries. Although only one person is responsible for writing these blogs, they are recognized in their communities. The research is done by others interested in the topic and then emailed to one "reporter." This more or less daily record of an industry helps to maintain a sense of historicity about a given topic though the blog archives.

Constructionism is a term coined by Seymour Papert of the MIT Media Lab, and is based on his work with Piaget in the psychology of childhood learning. Papert asserts that we learn by creating things and sharing them with others. Much of what is discussed in this report is predicated on notions of constructionism as a mode of learning beyond accreditation, where interest in the learning community, and situational interest (Hidi, 1998) is combined with subject based interest in a topic.

One of the most interesting examples of community based learning that goes unnoticed by most educators is the indexing of the Internet as presented in search engines. What is not widely known is that most of the Internet is not indexed by software spiders, but by people. http://searchengineshowdown.com/ how databases are generated and the relative value of them. The Open Directory Project (a.k.a. "DMOZ" http://www.dmoz.org) is the human indexing of the Internet. Dmoz is a type of service where armies of individuals search out and categorize the Internet in various hierarchies and topics. This act is similar to the great encyclopedic events of the past. Though realizing that it is impossible to categorize all of human knowledge, as the encyclopedias once hoped to complete, the act is one of a community trying to make sense of the infinite mass of undifferentiated information floating about cyberspace.

Some other initiatives are more specific, and tied to a fixed end-user environment. Librian.net is a personal and informational blog that attempts to contextualize information. Other sites, such as askme.com (or about.com), attempt to provide a forum for answering questions. Each sub-topic is hosted by an expert whose job it is to locate and evaluate all the information available on a specific topic.

(This section on MOOs is drawn verbatim from Nolan, 2001)

The MOO experience is probably the most important educational innovation over the past two decades (Bruckman, 1994; Curtis, 1992; Curtis, 1993). MOOs are a form of multi user simulated environment in which individuals create representations of people, places and things and share them with others. The emphasis is in the sharing with others (Nolan, 2001). Originally text only environments, with built in programming languages, mail, and chat, have grown into multi-media environments.

Online learning in polysynchronous environments is moving beyond the present "system of calculation," the substrate of our educational system, to one of "symbolism" and metaphor (Turkle, 1995). MOOs are environments in which text must stand in place of physical and conceptual things, and that which is imagined and described is created. And since there are no constrictions on what can be created in this manner, there is no de facto educational hierarchy in the MOO from which the educator derives power, though she may still control resources and access. For educators to function successfully in symbolic learning environments and within the non-hierarchical context of the Internet, they need to embrace the idea of learning through the co-creation of learning community, and then link these voluntary learning community spaces in informal setting to the formal setting, and in turn revolutionize it. At present, little research is available on how educators function in virtual learning settings (Nolan, 2001).

MOOs promote constructivist and collaborative learning (Bruckman, 1997; Papert, 1980). Users, be they educators, students, or casual users, work together and learn by building their knowledge as they build things. This social nature of MOOs is recognizable even in the development of the core database and server software. Within this context, the educator-student relationship is fluid, and educators must accept their place as students, and students will always be thrust into the position of having the opportunity to teach what they know.

MOOs are themselves valuable for two main reasons; their presence as the first generation of polysynchronous virtual learning environments, and the fact that they represent a direction for educational technology development potentially more rich and complex than existing environments can promise.

Polysynchronous communication is a term that came to my mind in response to comments that MOOs represented synchronous communication, in opposition to asynchronous technologies such as email, Usenet newsgroups, and learning environments such as WebCSILE, Participate, or Virtual University. People communicating polysynchronously not only talk synchronously (in real time) but also create temporal objects such as mail messages, newsgroup messages, as well as objects that can be experienced by others (Nolan, Kajihara and Lawrence, 1998; Davie and Nolan 1999). I refer to the communicative products as "objects" rather than as "texts" to focus attention on the object-oriented nature of MOOs. In MOOs, the text is contained within or mediated through "objects". These "objects" are programmed containers that allow the "text" to be manipulated by both the MOO server and the users of the MOO environment who create and interact with them. The contextualizing of the "text" within the "object" situates the communication more clearly within the technology, than it would appear to be otherwise. This form of virtual reality has potential as a dynamic learning environment. MOO spaces have been largely ignored by educational institutes for a variety of reasons. The main problem, for institutions and educators, is that they require a long term commitment to the environment, and are not amicable to short term, commercialized, learning experiences. The fact that control over the governance of the environment rests in the hands of the users, learners and educators, not in an administrative hierarchy, is seen as a threat to the sovereignty and governance prerogative of institutions. But they offer the most complete and immersive learning experience potentially available on the Internet.

There are a variety of alternative economic models for financing online learning that any assessment of the North American scene must encounter. Some sites, such as tutor.com take up the traditional role of tutoring online as a fee-for-service educational experience. Other sites use a combination of free basic content with an ancillary charge for value added content. The New York Times (http://www.nytimes.com) is an example of such a service that provides free access to new content, but charges for old articles ($2.50 USD for the ones we checked). Other services use ads, such as graphical (http://www.doubleclick.com) or text (http://groups.yahoo.com) banners to generate revenue for material which would be traditionally available only in print a traditional economic format. These environments do not present novel mechanisms for profit and cost-recovery, but there are models that have not existed before.

The rationale for discussing economic models in relation to an assessment of education is that the financing and governance of learning is probably the most problematic, regardless of whether the model is based on government funding or a for profit model, paying for education is a challenge on all fronts. In the future, successful online initiatives are those who will use alternative forms to finance initiatives and maintain revenue streams. The best models are those that are apparently transparent to the user, or provide a wide variety of mechanisms within a single environment, see Eudora below.

The most interesting of the new models is that of prosumerism. Prosumerism can be found in some of the largest corporations and technology initiatives. Linux, Redhat, Apache, personal web pages and open source initiatives are all prosumer spaces. The notion of prosumer, built on ideas originally presented in the works of Alvin Toffler (the author of Future Shock), is the idea that the online community is not divided into consumers and producers. But is a complex web of interactions of individuals and corporations who are at the same time consumers and producers. An individual surfs the web, and produces content on her web page that is consumed by others. A more complex example is found in the open source initiatives of Netscape (http://www.mozilla.org), Apple (http://www.apple.com), Apache (http://www.apache.org), RedHat/Linux (http://www.redhat.com). In these situations, major corporations provide free access to their source code. The programming community downloads the code, makes changes and modifications, identifies problems and security holes in the code, or creates modifications as they see fit, and shares their solutions with the larger community, including the corporations, who then select solutions that are incorporated into new versions of their products. Both the programmers and the corporations function as consumers and producers of the final product that is released to the general community as either free products or as commercial releases. This solution benefits all involved, as no individual could produce the complex code alone, and the corporations could not afford to pay for what they get from the community.

A more exciting development is found in the ideas of economic granularity (Teplovs,1995) and micropayments for access to information. In these models, one is working beyond normal fee-for-service or subscription economics. Micropayments, such as Derek Powazek's use of PayPal (http://www.powazek.com/, http://www.paypal.com/), allow individuals to make voluntary payments, as low as a fraction of a dollar, if an individual finds a service useful (See Jakob Nielsen, http://www.useit.com/alterbox/200110204.html; also https://secure.paypal.x.com/xclick/business=dmp%40fray.com&undefined_quantity=1&item_name=donation+to+powazek+productions&item_number=0001&amount=0.50&return=http%3A//www.powazek.com). Small personal projects become viable and profitable ventures when thousands of individuals pay a bit at a time, at a rate that is so small as to be unnoticeable to the pocketbook. Economic granularity envisions an automated system that carries this one step further. In this model, an individual’s account is credited every time their content is accessed, and a deduction is made every time they access information of another individual. Persons can provide content, offer courses, provide consulting for no apparent cost, but the more useful and valuable the content, the greater potential return for the individual. We have thousands of web pages of content, curriculum, research and information that has been regularly accessed by the Internet community over a decade, and an economic granularity model would make this work extremely profitable. For large corporations and educational institutions the overhead required to collect costs for specific courses could be relegated to an automatic system of checks and balances where valuable materials generate returns, and unuseful information does not. For users, there is strong interest in having a choice to pay, rather than being forced.

Aside from these interesting mechanisms of economics which would be central to any forward thinking educational environment, there are other models/mechanisms for content delivery that should be considered. The most obviously efficacious is that which is used by the distribution of Eudora (http://www.eudora.com). Eudora started as a free email program (http://www.eudora.com/presskit/backgrounder.html), and later spun off a commercial version with value added functionality, while retaining and maintaining the free version. With the advent of highspeed downloads, Eudora made the choice to combine the free and commercial product into a single release. Users download Eudora, and then decide if they want to pay for it, without having to download a new version. They merely pay the fee, and get the code to open their version. What makes the process truly innovative is that there is a third option that allows the user to get full functionality without cost as a way to preview the full version or to save money. Users can choose to accept advertising as a mechanism to get access to full functionality. They can also customize the ads they view to their areas of interest. The obvious advantage is that the user makes a decision, makes an explicit intentional act to participate is one of the three economic models. Mark Cannon's report (2000) states users would rather have banner ads than have to pay for content.

Other options that should be recognized are shareware, support-ware, freeware/postcardware, and group buying. Shareware simply enough is based on the honour system, and in a sense is not much different than alumni support for educational institutions. One version of shareware is disable-ware in which functions are disabled until payment has been made, but this idea is not as good as the combination-ware of the Eudora example. This could be valuable in education, however, where functions of testing and evaluation and accreditation could be withheld, but the content could be freely available. Support-ware extends this notion one step further, and is useful in terms of educational models. The content/code/software/environment is free, but support and customization would be at a significant cost. The courseware and everything around it could be free, but all costs of dealing with an individual, or support databases, would be the revenue generating mechanism. Freeware/postcardware are similar to shareware but are either free to the user or request that a postcard (or similar item) be sent to the vendor. Group buying is still product focused, however, the product is supplied only buyers when a critical mass of interested buyers is obtained. This could be useful in offering courses based on interest instead of traditional semester timelines.

Archiving, as it relates to this report is an issue of choices. Knowing what content requires archiving, where and how to shape content, and issues of knowledge mapping are all important to online learning (http://public-cmaps.coginst.uwf.edu/cmaps/). As well, the archiving of accurate or correct content is an issue. As with traditional learning into questions of valid information, students study lies/untruths until experiments are done to demonstrate to the student how something actually works. Information that is brought to the learning environment must be adequate otherwise only ignorance is shared (Harasim, 1998; Pettit, 1998). The ability to discern the validity of archived knowledge is central to online learning being accepted as equal to other modes of learning. Putting date stamps on pages that are server generated is one step towards better archiving. Another idea would be to provide an external referee "lock" symbol to promise a create date or constancy of information. This lock could be similar to the SSL visual lock now available for ecommerce, banking and other secure transactions.

The notion of how data is maintained online is crucial for the credibility of online learning. And at present, the state of affairs is horrific. The Internet is totally lacking in any sense of historicity, functioning in a continual present, as far as the majority of users are concerned. A web page can change in a day, with no sense of what was there before. No other medium has had such a transient face. Very little has been done to counter this state of affairs, because for information architects and programmers, historicity is not an issue of concern. But for the Internet to be a credible learning environment, it must generate some sense of historical contextualization of its parts such that it is possible to discern changes over time. A continual present state of affairs has never boded well for any educational institution. The Sorbone, Harvard or Oxford gain their stature and credibility as a result of their existence over an extended period of time that is easily apprehensible by the average individual. And it is more than affixing a "since 1400" marker on a web page.

Educational institutions’ value is contained in its archives of historical, academic and research accomplishments which are used and presented to the public. Oxford’s Bodliean library has opened its images to the world, Encyclopedia Britannica has gone online, but this placement of content does not get at the problem. The problem is one of historicity of content. And mechanisms must be in place for individuals to look backward in time to see what has gone before it. Imagine a company or institution doing what I do in my course, allowing students to view previous year’s web pages, content and assignments from 1997-2001. And think of what it means for the credibility of a course that has been taught online for 4 years when prospective students can see who it has grown and developed and gained in value and sophistication. Student still has to believe that previous years weren't modified.

From a marketing stand-point historicity is eschewed, showing only the biggest and the best. But this is not the case for knowledge, information, learning and education. And should not be the case for the Internet. I think many would be shocked to see what the Internet looked like in 1992. Ed Kroll’s famous book The Whole Internet User's Guide & Catalog which led the way for the how-to craze of Internet books. It is a completely different planet from what it was then, and Kroll will go down in history as the individual who saved the history of the Internet. The value of the net is not predicated on the failed IPOs of yesterday, or the hype of tomorrow. It is predicated on the understanding of what it is now as a result of what it has been in the past. And it behooves any and all organizations to realize this when it comes to information and learning, especially when it is anathema to corporations.

These notions of authenticity and the re-writing of history are important, not just for education and accreditation (noted below) but also for the self preservation of online organizations. Without the ability to contextualize what has gone before within future plans, the tendency is stagnation in a continual present that can neither remember its past or envision its future.

Ways in which historicity is being maintained is by personal archives on various topics that are Archiving the Net. The Lost Library of MOO (http://www.hayseed.net/MOO/) is a great example of this sort of archiving environment. It is a repository for information that would otherwise become lost. The library itself has moved more than once, being passed down to various new librarians. Those of us who have nuggets that have become lost on other sites have become sub-archivists with contributing sites full of papers and archived web sites of great importance that would otherwise have become lost when a server crashed, or a student graduated and their web space was removed.

Traditional archives have also found their way onto the internet. Sometimes they are used as a traditional archive: where users can search for information and find a digital representation of that artifact. Oxford's Bodliean library (http://www.bodley.ox.ac.uk/dept/scwmss/wmss/medieval/browse.htm) has scanned its images at high resolution, allowing users to look through the archive regardless of where they are. This digital archive helps give access to those who might not be able to physically access the library. For those who do plan on viewing the collection it allows for initial research time with the digital version of the documents.

Libraries are not the only kind of institution that have taken advantage of the internet to present their archives. While the Bodliean library has placed still images on the web, other collections have used different technologies to present their artifacts. The San Diego Zoo and National Geographic have both placed bear-cams on the internet. The San Diego Zoo Panda-Cam (http://www.sandiegozoo.org/special/pandas/pandacam/index.html) allowed users to watch the development of a new-born panda in an area that had no public access. Footage is archived on the site for users who have not been watching since its inception, providing a video archive of the development of an endangered species. National Geographic has also established an archive of an impossible place: The Bear-Cam in Alaska (http://www.nationalgeographic.com/bearcam/). Their camera is at McNeil Falls, Alaska and is the, "site of the world's largest gathering of bears." A brief introduction on the site tells users at what time of day they are most likely to see the bears -- the camera is only on during the appropriate season during the year. Video footage from National Geographic's library gives users examples of brown bear behaviour they are likely to see. National Geographic has since added Wildcam features: elephant seals, otters, polar bears and stellar sea lions.

Traditional archives are also available for new and novel uses on the internet. It was inevitable that users would eventually be able to search their local libraries and reserve books online; however, some collections are not being searched with the intention of ever finding a paper copy of the book. Rather, they are being used to research the existence of a book. The Library of Congress (http://www.loc.gov), although 'local' to some, is now being used on the internet to find bibliographical information about a given topic.

Practitioners of technical design tends to throw around terms like backend, frontend, bandwidth and infrastructure. What we’re interested in is what this means in terms of accessibility and learning. There are many different kinds of things that impact the technical design of a project. The easiest division is hardware/software. There are three facets of these divisions: server-side, client-side and in-between. In any given interaction on the internet, the user will have access to a variety of programs on their computer. One of these programs will be used to call up the "outside world." Although one can be led into believing that our computer communicates directly with another when we call up a webpage, this is not the case. The user's request for information is first filtered through a series of other computers and then the response is filtered back. This negotiation through a network of computer servers cannot be easily modified but can result in a "slow network." The terms most often thrown around for the slow network are lag and bandwidth. While some companies try to reduce the download time of own products, others are looking at what can be done is there is an unlimited amount of bandwidth. Professor Nancy Paterson is currently developing a VRML library at Seneca College (http://www.bccc.com/nancy; library: http://www.thelibrary2.com/; content: http://www.bccc.com/nancy/6dos.hmtl; plug-in required: http://www.cai.com/cosmo). Students on the Seneca intranet can experience the Canadian Parliamentary library in a 3D environment. Like Snow Crash's library, users are able to pull up information from the web (e.g. environmental forecasts) through the library interface (Stephenson, 1992). While a striking example of what can be modeled in a perfect scenario, the level of complexity that the VRML environment (3D) allows for is simply not required to accomplish user goals on the internet, not to mention impossible given current bandwidth limitations. As bandwidth increases there may also be an increase in perceived need for such environments.

In many cases the user may choose to access certain material in a preferred format depending on what they are going to do with the content. The Gutenberg Project (http://www.gutenberg.net) has a cast of volunteers that transcribe, proof read and publish texts on the internet. While they are nice to have as a reference, many prefer to read these texts in paper format. The reasons why users prefer one format over the other range from ease of transport/portability of paper to eyestrain and legibility and finally cost (borrowing from the library vs. internet access vs. purchasing a paper copy). That doesn't mean the Gutenberg Project is without value though. It may simply mean that readers prefer to enjoy fiction away from their computer. A scenario would be to use the Gutenberg Project to print out specific chapters of a political science text that was part of a required reading package. The user could then customize the font and other print options of pages wanted printed and read the rest online. Giving the user the control that has been proposed at a larger scale by the publishing world for print-on-demand publications.

While users may decide how they want to access information, there are barriers that will prevent a perfect match of what the server-side can deliver and what the user requests and is capable of handling. Decisions have to be made about what kind of software they end user will have. Real-time (synchronous) support can be delivered through a variety of software solutions; however, each will have their drawbacks. ICQ (http://www.icq.com) and AIM (http://www.aol.com), two of the most popular instant messaging packages, are still software packages that a user must commit to downloading and installing. A solution that is simply not possible on an individual basis at large, academic institutions, not to mention at home where there is likely little to no technical support. For those using a computer that is at most 4 years old with a browser that is at most 2-3 years old, the easiest synchronous communication tool for the user to 'install' is a Java-based environment called up through a web browser. There are still a number of limitations in using it. For example, a Java applet running in a web browser cannot be resized; forcing the user into a screen of one size. Other applications like ICQ, AIM and telnet allow greater degree of control on the working application (e.g. resizing, adding and/or skinning while still connected).

For some of the technical decisions that are made the user is unlikely to ever notice the difference in the service that is offered. The decision to use a Mac, Windows or unix/linux web server is not likely to be affected by who the end user is, but rather: the availability of programmers and program languages/ software availability for that platform. The user will care whether or not the service is fast, but they won't care if it is slow because Perl was used instead of flat HTML files. Indeed some studies have shown that if content is "good" users will perceive a service as faster even if it is not (http://www.useit.com/alertbox/9703a.html and Spool 2001). Even though the user may not noticed the soft or hardware used for a given site, the ability to grow a site's features and ability to handle increased traffic may be affected by the initial choices that are made.

In some cases the hard/software used may be an important part of the integrity of the service being delivered. Whether that be matching corporate sponsors with the software/hardware they provide or matching a political philosophy associated with a given software/hardware configuration. For example: many of the open source support forums use only open source software. These are all questions that speak to the design of educational environment, questions that are almost always given secondary consideration to what is considered to be the special or pedagogically valuable idea that the environment is trying to express.

Elearning has an exceptional hurdle to overcome. It must face the challenge of a new way of communicating hat requires a different way of presenting materials. Many academic institutions have purchased WebCT (http://www.webct.com) and Blackboard (http://www.blackboard.com) licenses. Both of these packages allow an integrated web space where instructors can easily upload course materials, grades can be recorded and students and teachers can interact synchronously. By promoting the ease of course dumping onto the internet, WebCT and Blackboard have promoted the current generation of teachers into thinking they are able to teach online without any modifications to their content or style of delivery. While some students may be able to excel in an environment where teachers simply dump information on the internet, it is despite of the environment, not because of it. Teachers who have recognized they must learn new technologies to take advantage on online learning become frustrated in the increased amount of time required to learn new technologies (New York Times, 22 January, 2001).

Where WebCT and Blackboard may excel is as learner management tools, or more appropriately, administrative assistants. ROSI at the University of Toronto (http://www.rosi.utoronto.ca/) currently performs a limited learner management role. Students are able to register for courses, check outstanding tuition owed, request an unofficial transcript, etc. It performs a limited number of tasks and does so well. Specialized tools for specialized tasks, accessible from the web, is a preferable choice to any present or imaginable all-in-one package. By promoting the best solution for each part of elearning, teachers may be encouraged to learn more about delivery of content instead of trying to dump their courses on the internet and expect results. They may also be able to spend less time learning how to use the technology if they are given the best technologies to use instead of all-in-one packages that might not excel at any one thing.

The technology that is used may be influenced by the ability of the user to use a given piece of software. Ted Clark at the University of Melbourne (http://as1.ipfw.edu/2000tohe/papers/clark.htm; if password protected use 2000 and tohe) used animated gifs and audio clips to illustrate to students how to install and use specific software. Although these clips may take longer to download than text-only descriptions, they require no plug-ins. Server logs showed Clark that when given the option of text-only vs. animation, the user opted for animation more often than text-only. However, for users who may be relatively new to the internet, the task of downloading, installing and configuring a plug-in can be incredibly frustrating, not to mention impossible in many school networked environments. Using a plug-in-free method of delivery allows a greater percentage of successful page views.

When visual and audio clips are provided, the content provider must be aware that there are (at least) two categories of people unable to access their information: the disabled and the super-enabled. The W3C (http:// www.w3.org) has developed guidelines on how to ensure all users are able to access content. The WAI (http://www.w3.org/wai) has developed a checklist that highlights alternative modes of delivery for non-text information. They recommend long descriptions for images and transcripts where appropriate. These accessible formats allow users with low-end/slow connections to access information without having to wait long periods to download clips. It also allows users with physical and or learning disabilities to experience the information through alternative delivery mechanisms (e.g. blind users who have Jaws installed (http://www.hj.com/JAWS/JAWS.html)).

At the other end of the spectrum are the super-enabled users with handheld devices instead of lap/desktop computers. Services like Avant Go (http://www.avantgo.com) allow users to download content from sites to be read offline later. There is limited visual support on these devices and in most cases no audio. Some sites have taken advantage of the new format (e.g. MapQuest’s (http://www.mapquest.com) download a custom map that you make on their website); however, sites using plug-in dependent technologies on their site may eventually be required to maintain a separate site for those using PDAs/handheld devices to read the internet. Stylesheets allow some control, but for text-only pages. As of right now most developers using plug-ins are not using databases for their content within their movies. When content is stored in a database re-purposing the content into an HTML version would be almost trivial.

Those surfing the internet on lap or desktop computers continue to pose a challenge for developers who wish to create sites that will work across browsers (Netscape, Internet Explorer and Opera being the most common right now but WebTV browsers have started to show up in TVOntario (http://www.tvo.org) usage stats) and across platforms (Windows, Unix, Mac being the most common). While sites targeting a general audience can take some liberties in how much testing they do on each of the combinations mentioned above, there is no reason to eliminate potential users simply because of their kind of computer setup. (As of Issue 99, A List Apart (http://www.alistapart.com/) has chosen not to support non-W3C compliant browsers. Although content can be viewed on their site with older browsers it looks like crap.) As Jakob Nielsen (http://www.useit.com/alertbox/20010204.html) stated in a recent Alert Box that just because a user may be stupid now, think of all the people you may end up ruling out later as more and more people get connected to the internet.

Having an accessible site has become increasingly important for the US market with the adoption of the American Disabilities Act 1990 with specific requirements for accessibility (http://hotwired.lycos.com/webmonkey/98/21/index2a.html for more info). On the internet, content is king. Making sure that the technical set-up does not prevent users from accessing a site is paramount. Technical solutions should be used to enhance usability of the site, only when completely necessary (and if you happen to be a design zine like K10K, technical solutions are completely necessary). Even Macromedia has come to realize exactly how much users hate Flash. Build for today's needs with forward planning, but do not overbuild. A working model of the good, bad and ugly will prove more useful to a re-development team than no model at all. This philosophy is as important for design as it is for content of learning environments.

The catch-phrase "content is king" has certainly not escaped the internet. Although we are not calling it, "the information super highway" anymore, people expect to find what they are looking for. A recent study of Ontario internet users states that one quarter of users return to a site for the content. And nearly 20% will not return to a website if they found it difficult to use. It's not enough to offer a great buy on your site and throw up a few banner ads. You must have content; it must be easy to find within your site; an it must be good. There are a number of books that have been written about usability and information design for the web. Most give a clear message: tell the user where they are; help them find their way around by giving them clues; and keep your writing concise. While many of the rules are general, there will always be exceptions to anything that's written down about the web (usability testing with an average user vs. there is no average user). Instead of giving rules about information design and dropping a few URLs that are already highlighted as good examples of design, it is better to highlight issues that get less of the limelight. It should be noted that these are not new (or even novel) ideas for many of the texts on information architecture -- they are simply issues that are less likely to hit the popular media anytime soon. And which are crucial for successful implementations in the future.

A friend of mine jokes that she is my "average user." Whenever there is a need to test out a concept whether it be for navigation aids or program instructions, she is told to take a look. While it is wonderful to have a sounding board, it is always important to remember: there is no average user. Each person coming to a site will have a different motivation. For some this could be the distinction of research/reading vs. entertainment (http://www.alistapart.com/stories/who/), others might be coming at the site with the distinctions of learning styles (visual, textual, experiential, prior knowledge) (http://www.websavvy-access.org/resources/wai_newgl.html); and still others might be searching for an item, by any number of methods (Information Architecture for the World Wide Web).

Instead of limiting users to one system of navigation, provide them with several different ways of navigating. Use images and text; provide site maps and searches (more the better if the search engine also provides a spell check and suggests 'alternative' spelling for what you entered). The goal is not to overwhelm the user with too many options, but rather to allow a user to focus in on what works best for them. One option is to determine what a user's surfing style is and then give only those navigational aids (Teplovs, 1997); however, changing a website's navigation system (even if under or un-used) is considered bad form.

Different styles within the same site are more common in sites that have feature articles that have been (creatively) designed. Even within different presentations of information, there should be a consistency to help guide the user. For example: the fray (www.fray.com) uses the same format for each of its stories: first select a story, then experience the story (includes finding links to move forward in a linear story!); then read responses and add your own if you choose. In contrast, Kairos (http://english.ttu.edu/kairos/) provides an inconsistency of interaction in each of the levels in their site. While their "home" page is clearly grouped, uses a white background, tables for layout and a lot of negative space, its articles are framed and often difficult to navigate. On more than one occasion, frustration with experience with the text has led to abandoning the article in search of the same content in a format consistent with the "home" page. As Northrop Frye stated: I know what to expect from Agatha Christie, even if I don't know "who done it." This is not to say that I want a printer-friendly version of all websites, rather, I want a consistency of interaction from individual sites. Kalliber 10000 (www.k10k.com) uses enough scripting to choke a horse. It takes "forever" to download and can only be used with the latest versions of browsers, but it is consistently high in its demands on the user. Both in terms of technology and patience to filter through the text in too small windows that show only a few lines and animated buttons with no purpose. It is also updated regularly and offers novel ideas that can be applied on more accessible sites with a little imagination. It is not necessarily the quality of the design or the content of the site that distinguish these two sites, but rather the quality of the experience.

National Geographic (www.nationalgeographic.com), PBS (www.pbs.org) and the Discovery Channel (www.discovery.ca) have all brought their experience in producing documentaries onto the web. Although the Discovery Channel and PBS provide a less engaging front page than National Geographic's all three have important content, feature articles that engage the user. Because of their use of design, images and ease of use, these educational content producers are able to capture the user in content they might not have otherwise been interested in. Situational interest is key as a motivating tool for learning. The feature articles/ documentaries available on these sites are not complete learning units. Rather they can serve to motivate users into continuing their search for more information about them.

There are almost an infinite number of methods and directions to access information on the internet. The primary mechanism for actively searching for information on the world wide web is through the search engine. Mark Cannon's study puts search engine searches at the top of how users find websites (Cannon, 2000). Over the past 5 years there have been important changes in terms of the best engines out there. From Yahoo, to Altavista to Lycos to Hotbot (now combined) to the current #1: Google (http://www.google.com) each new technology reaches ascendancy. A no-fuss-no-muss option, Google gives indexing without all the banner ads and categories on the front search page. The issue is accessing content, to repeat this fact over and over again. But not just raw content. Presently search engines are useful for all levels of users, while specific needs and newer users rely on portals which tailor content to specific interests. The goal is to tailor sites to specific search styles and accessing information in a manner that takes up cultural, linguistic, bioregional differences, learning styles and physical disabilities. When you consider the rapid growth of the internet (http://www.nua.ie/surveys/how_many_online/index.html) in many different contexts, over the past six years, the requirement of increasing diversity of users in the North American learning contexts

When considering issues surrounding the visualization and presentation of information, there are a few innovations that are worth considering from an educational perspectives aside from what one might encounter in the various *ML (markup languages) and wc3.org initiatives, particularly cascading language styles and skinning. Simply put skinning is the ability of the user to completely reconfigure the look and feel of a site. When a site has been designed with a default look and feel, skins allow a user to chose from alternative design options, or to create their own. There are no examples of skinning in educational contexts, and none have been identified in web contexts, but there are many examples of skinning used in consumer applications and utilities. SoundJam, the MP3 application (http://www.soundjam.com/) is a good example of skinning and the community around it. The educational value of skinning is obvious. Allowing users to configure and recontextualize, control, their own learning environment as a mechanism for promoting situational interest and promoting motivation for the context for learning.

Cascading Language Sheets are a conceptual extension of wc3.org’s Cascading Style Sheets, and something that Project Achieve (http://achieve.utoronto.ca) is planning to implement in their next generation CVLE (Collaborative Virtual Learning Environment). The idea is that most sites online are mono-lingual, or sometimes bilingual; http://achieve.utoronto.ca/sadako is a bilingual English/Japanese site. However, Project Achieve is a polylingual environment, and has had to struggle with experimentations communicating in Tiawanese, Icelandic, French, English, Japanese from room to room in the virtual spaces.

Limitations of old technology meant that every room had to be specifically coded, and user’s web browsers needed to be configured, and Java applets modified to facilitate different encodings such as ASCII, extended ASCII, sjis, big-5 for English, European languages, Japanese and Tiawanese respectively. Note that these are all languages that North Americans wished to communicate in at different times. Though Project Achieve successfully modeled the ability to maintain a user configurable polylingual environment, the experience highlighted the need for a more sophisticated solution. This solution takes the form of Cascading Language Sheets (CLS), and has been designed into the next generation CVLE design.

CLS is predicated on the notion that the user must negotiate a common language with other users and with the environment in which the user is communicating. A variation of this exists within some web browsers where users can chose encodings they feel comfortable viewing web pages in, but CLS takes things one step further. CLS allows for a matrix of language and content where each actor (player, user, object) contains language data content or usage information in the languages in which she feels comfortable communicating. For example we have a user named Sarah entering some virtual rooms named A, B and C. Sarah understands English and Japanese, but can read a bit of French. She has her CLS settings set to English/Japanese/French, meaning that this is the order she would prefer to see information, if it is available. When Sarah enters room A, the room scans Sarah and realizes that she prefers English/Japanese/French, but it only has Japanese/Russian as options, so it presents data to her in Japanese. Room B has information in English/Japanese/French, and presents data in English, her first choice. Room C has none English/Japanese/French as options, but only Urdu. Having no common language, the environment in which Sarah and the rooms are located intervenes in Sarah’s first language asking her if she would like the contents of the room translated, using a commercial web-based translation site (http://world.altavista.com/ and http://www.systransoft.com/Homepage.html), into a language she can understand. Another option, when the user does not have the fonts/technology to read the contents is Ka Ping Yee’s Shodouka (http://www.shodouka.com/). Shodouka translates Japanese web pages into images that can be read/viewed by anyone even without a Japanese language computing environment.

Unicode, (www.unicode.org), is a major part of this solution. Though Unicode is only a 16 bit code (ASCII is 7 and extended ASCII is 8), it does not have enough space to contain all characters. Chinese has over 60k characters, and to provide access to them all would require 32 bit character code.

Aside from the obvious value of this tool pedagogically, it has enormous currency in countering the prevailing Englishization of the net in mainstream communication, but it also speaks to the same notion that skinning addresses. This notion is that online learning must conform to the needs of users and take every opportunity to not stump the user because of some limitation or incapacity for the user to interact with the data.

Education is not about content or mechanisms of delivery or even about techniques of summative evaluation. These are all ancillary to the twin acts of teaching and learning. Learning is about intentionality, task dedication, situational interest and motivation. The locations of interaction, whether polysynchronous, or merely synchronous/asynchronous, are valuable as learning environments insofar as they are able to encounter the user on the user’s own terms, and minimize the need for the learning/user to reconfigure herself to the environment.

The bulk of the contents of this report directs attention to non-standard technologies or technologies and practices. This has been an intentional positioning, based on the opinion that what is presented as educational technology and learning environments are predicated on out moded innovations and learning models that preceded the explosive growth of the world wide web and the concomitant increate in computing power and bandwidth.

Nolan, J. (2001). The Techneducator Effect: Colliding Technology and Education in the Conceptualization of Virtual Learning Environments. Unpublished PhD dissertation. [Web Page] [http://achieve.utoronto.ca/jason/dissertation.PDF]