ACCESS

The Brooklyn College Library
& Office of Academic Information Technologies

Volume 6                       Spring 1997

Meet Nick Irons: Academic Computing's Latest Installation

Distance Learning: Some Notes on Academic Issues
           Barbra Buckner Higginbotham

Windows 95: Multimedia's Best Friend - The Best Operating System for Multimedia
           David Ditillo

: - ) When You Grade That: Using E-Mail and the Network in Programming Courses
           David M. Arnow

Government Information on the World Wide Web
           Jane Cramer

Spring Research Tips: Pressed for Time to Do Research?
           Miriam Deutch

Learning is proximal. An originator of information (the teacher) is in close, proximal contact with a receiver of information (the student). Information is conveyed visually and verbally from the teacher to the student in a linear manner and in real time. Physical distance between the student and the teacher is limited by the reach of the human voice and by the range of human vision. 

With the invention of writing in Sumeria about 10,000 years ago, it became possible to transmit knowledge over a distance. The written word was not only permanent, it could be delivered to the student at a distance from the originator. "Distance learning" had begun, and the originator of the information (the teacher) could now be separated in either time or space (or both) from the receiver (the student). In all these 10,000 years, however, no textbook has ever been able to replace completely the powerful interaction of proximal contact between a teacher and a student. 

New technologies are now rapidly eliminating this ancient restriction. It is becoming possible to retain the teacher/student relationship in learning, even while the two parties are separated by great distances. Since information must still flow, however, communication between parties must make use of a technological intermediate. It is the nature and structure of this intermediate that defines and limits the possibilities for distance learning. 

Any implementation of distance learning must first appreciate the opportunities and limitations of these technologies. Given a need for distance learning, and given a body of information to be conveyed, the educator must balance the pedagogy with the best and most effective methods of transmittal. 

Flow of Information

Information flows between parties in a distance learning setting. This information must be converted from its original form, travel along or through some medium, and then be decoded back into its original form at the end of its travels. At each of these three stages, the educator must be aware of the basic limitations and advantages of the various alternatives.

Encoding and Decoding: A Golden Rule

"When converting between types of analog information, always take the shortest route." Never break this rule. It is a property of the second law of thermodynamics that "things always get messier."

A tiny degradation occurs every time a process is performed. So, when the human voice is converted from analog sound waves to analog electric amplitude waves, some information is lost. Sound quality degenerates and subtle nuances vanish. For example, as a microphone converts vocal tones into agitated electrons, not everything is retained. At the other end, as the electric current is decoded back into sound waves by the loudspeaker or headphones, once again sharp peaks of amplitude are softened and timbre is sacrificed. Noise is introduced and the voice coming from the amplifier and speaker has nowhere near the original quality. 

If a voice is first recorded onto tape, the tape played down the wires, the voice re-recorded at the other end, and then the second tape played back, the degradation will be even worse. The longer the "route" and the more conversions necessary, the worse the final product. Even the professional broadcasting industry, using the best possible techniques, are plagued by this property of quality loss. They spend large parts of their budgets on high-end analog encoders and decoders and are turning more and more to digital devices where the degradation can be held to a minimum--but even with all this, it is impossible to prevent some degree of loss. Thus, when preparing for distance learning, the educator should always plan the "shortest route" between the teacher and the student, and use digital equipment whenever possible. 

All types of analog information, not just sound, are affected by the number of times they are encoded and decoded. Pictorial imagery is especially sensitive and its quality degrades even faster than sound, even when digital technology is used. Although there are many ways to reduce quality degradation, it is a good rule to "Never break the golden rule." 

Bandwidth

Along with the quality of information being transmitted, the educator also needs to consider the quantity of information that is passed. During a phone call, for example, information travels along the wires as a linear sequence of words. Ignoring the technical details of how this happens, the wires must be capable of transferring one "bit" of information at a time, one after another, rapidly enough so that the decoding device at the other end is never kept waiting. Normally the capacity of the phone wires is so high that passing single "bit" information (a phone conversation) is never a problem. 

Passing a television picture along a wire takes place in a similar way, but the quantity of information being passed is enormously higher. To begin with, each tiny dot on the TV screen represents three "bits" of information (chroma, luminance, and synchronization). If you include sound, four "bits" of information must pass more or less at the same time along the wire. As the wire can only pass one "bit" at a time, the four "bits" are lined up and sent one after another, a bit like four conversations going on at the same time down the same wire. 

An average picture on a TV monitor consists of 172,800 dots, which (ignoring the sound) needs 518,400 "bits" of information before it will appear. A TV picture screen changes 30 times a second, so, for one second's worth of information to appear, the wire must carry 15,552,000 "bits" of information. No telephone wire has the capacity to carry all this information without serious delays. This is why you cannot get a TV channel over the telephone wires and why TV Cable Companies have to provide special, high capacity wires. 

This capacity of any transmittal medium to carry information is called "bandwidth." As decisions are made about the modes to be used in distance learning, bandwidth is a critical item of consideration and budget. As the bandwidth widens, the costs escalate astronomically.

One-Way or Two-Way?

Most telephone conversations are two-way. Parties at both ends of the wire can transmit and receive information. Television broadcasts are one-way. They send, we receive. In a classroom, a good teacher engages in two-way learning. It is important that students be able to engage the teacher and carry out a dialogue. This is one reason why even the best books or, these days, just a video tape, are not enough. Instruction and learning are two-way processes. When setting up a distance learning protocol, therefore, the two-way nature of learning must be taken into account. 

If the budget permits, a two-way, real-time engagement is the best. But, as pointed out above, two-way information transfer where both parties are using "live" TV and sound, would require that the bandwidth of the transmittal medium be twice that of the system that currently brings cable TV. This is possible, but expensive. At the present time, distance learning protocols compromise. High bandwidth may be used for the signal going from the teacher to the student, and a much narrower bandwidth used in going from the student to the teacher. 

Planning the Product

Before any distance learning can take place, the educator must make three important sets of decisions; the learning mode(s), the delivery system, and the bandwidth of the transmittal medium. All of these decisions are interrelated, and choices made in one area obviously influence the range of choices made in another. Some choices are decided by budget. As discussed above, high bandwidth media are very expensive, so if the budget will not go to these stratospheric levels, full two-way video conferencing may not be an option. Finances may fix the bandwidth, but the educator does have options in the other areas. 

Linear vs. Nonlinear Pedagogy

While the range of pedagogical styles is beyond the scope of this analysis, broadly speaking, distance learning will normally involve some mixture of linear and nonlinear pedagogy. Because of the unique features of distance learning, however, it is important to decide on this mix early in the planning stage. 

Linear. At one extreme of teaching style is the totally linear presentation of the syllabus. This mode of information presentation is seen most clearly, perhaps, in disciplines such as chemistry. With some variation, most courses begin with the principles of atomic structure and work their way in a linear fashion up to the properties of complex molecules. Information must be presented in a linear sequence because later facts and ideas depend on an understanding of facts and ideas presented earlier. Talking about the structure of benzene before the principles of aromatic chemistry or the nature of the carbon atom would be a pedagogical nightmare. 

This mode of presentation is similar in some ways to watching a television drama. Packets of information arrive in a linear sequence, and the theme or story line develops as later packets of information build on what has gone before. Tune in at a point midway through the drama and you have no idea which character is which or why one of them is about to jump off a roof! 

Nonlinear. At the other extreme of teaching styles is the totally nonlinear approach to presenting a body of knowledge. In this mode, information is presented as discrete packets that are essentially self-contained. Each packet of information is important in and of itself, but as more and more packets are added to the mix, the whole becomes more than the sum of its parts. 

As each packet of information is self-contained, the sequence of packets is not critical. The student, therefore, has more freedom to construct the flow of the syllabus according to her or his learning abilities and interests. Depending on how each packet is constructed, it is also possible for the student to engage in more interactivity with the information. Within any given time constraints, it might also be possible to revisit certain packets of information later and reanalyze their content in the light of later discoveries. Although difficult to construct in its more extreme forms, certain elements of nonlinear education are appearing more and more in older traditionally linear curricula. 

Continuing the analogy of television, nonlinear education is similar to a series of cooking programs. Each episode teaches the student how to prepare a certain dish or learn a certain technique. Assuming these episodes are on videotape and may be viewed in any order, one student may begin a cooking course learning how to make sauces, go onto vegetable preparation, and finish with how to manage roast meats. Another may start with the meat episodes and end with vegetables. At the end of the course, however, all the students are accomplished chefs, although they all took different routes to get there. 

Distance learning can take advantage of the best of both these styles of teaching. Where the lack of a high bandwidth medium may restrict a two-way, linear style of teaching, switching to a nonlinear mode for at least part of a course can solve two problems at once: high cost and low interactivity. 

Time-Dependent and Time-Independent Modes

Distance learning is either time-dependent or time-independent. A class that meets once a week in an agreed time slot for one hour is time-dependent. All participants must gather at the same time for the length of the class. A takehome exam is time-independent; the student can take the exam any time he or she likes. Other parts of an education can also be time-independent. A textbook, for example, is time-independent; students can read the text any time they like. With distance learning, all the tools of education and multimedia combine to enrich both the time-dependent and time-independent modes. 

Real-Time Functional 

Traditional (Wide Bandwidth): If expense is no object then full bandwidth, two-way audio and visual communication has a lot of advantages when traditional education techniques are converted into a distance learning mode. At Maryland State, a network has been constructed with the help of $30 million from Bell Atlantic which uses full two-way audio and video signals. The state hopes to have 300 schools, colleges, and museums linked to this network in three years. This type of system essentially creates a second "classroom" at the distance site. Video images and full audio of the teacher are delivered to the second site, and the teacher can receive full video and audio images of the distant students. In Maryland, Bell Atlantic is donating $50,000 worth of equipment to each of the sites and to help them hook up to the network. 

All the imagery needed for classroom instruction is created in real time (even the blackboard, if necessary), and all these signals are transmitted along high bandwidth cables. As discussed below, linked cable distribution systems have some advantages and some disadvantages. In North Carolina, for example an advanced network is being created which not only carries full voice and video signals, but is also capable of high speed data transmission. As technology changes, data transmission will undoubtedly play a greater and greater role in distance learning, and some networks are already trying to take future upgrades into account during current planning. 

NonTraditional (Multimedia-Dependent): Pre-created imagery considerably enhances educational pedagogy and is certainly applicable both to proximal and distance learning. Known today as "multimedia," these pre-created materials can be computer generated, built "on the fly" from existing computer programs, analyzed using computer simulations, and captured on and from videotapes, databases, or the Internet. The choice and range of these materials, commercial or self-generated, is staggering and growing by the day. 

Not many schools can go as far as Notre Dame in providing these multimedia services to its faculty. At this institution, DeBartolo Hall has become the heart of what is called Media-on-Call. Each room is connected to the main network by four (!) high-speed, high-bandwidth, fiber optic cables. Two of these are used for full two-way communication .and two are reserved for use by not yet invented technologies. Along these cables flow all possible types of multimedia. Professors in every room have access to six audio or video sources and during classes a simple push of a button brings up pre-selected materials on large-screen projection units or stereo speakers. All classrooms can receive the Internet and satellite signals, and four rooms can originate two-way audio-video teleconferencing for distance learning. 

Educators at Notre Dame are already claiming improved results from their $25.5 million system, but institutions lacking Edward J. DeBartolo's $30 million dollar generosity are, nevertheless, trying to offer decentralized distance learning. Just one year old, the Virtual Online University offers 350 courses in eight different schools. The system uses MOO, multi-user dimensional and object oriented software of a type familiar to computer game players. Narrower in bandwidth, these kinds of approaches can reach larger audiences. Anyone with a computer and the course fees can sign up and log on. Once in the "Virtual Campus" a click on an object brings a student into a "classroom," enables her or him to meet the teacher, allows visits to other "buildings," or allows a student to take a seminar and record every bit of instruction for later study. 

At the University of Cincinnati the bandwidth is even smaller. John G. Bryan no longer attends a classroom or gives lectures. Although he is teaching technical writing, he has yet to meet one of his students face to face. Mr. Bryan uses "Daedalus Interchange" conferencing software and simple IBM computers to create a virtual classroom for his 21 students. Although he has yet to solve all the technical and human problems of distance learning using this system, and he is not yet convinced he has found the right formula, he is looking into ways of eliminating the "classroom" concept and taking learning further into virtual reality. 

These three examples show that bandwidth is important to distance learning. For those on limited budgets, a hybrid system may be the best solution. A course could be booked into a full teleconferencing classroom for one meeting a month (full bandwidth), meet three times a month on a "Ce-You-Ce-Me" conferencing system, and exchange tests and answers using e-mail. (A "Ce-You-Ce-Me" system uses a $90 camera, a personal computer, and "one-bit" data transmission to send images between computers.) 

Time-Independent

With time-independent distance learning other factors enter the equation. Standford University is testing the concept of "education on demand." Supported by a $500,000 grant from the Alfred P. Sloan Foundation, two classes in 1995 were taught in a time-independent manner. Engineering students were logged on to high-speed networks which enabled them to download audio and video lectures, text and graphics. Called ADEPT, this approach builds on the idea of pre-creating a full range of imagery from videotaped lectures, static images, text and other materials needed for a course and then providing the technology to access these images at any point in time. This is where those "other factors" enter the equation. Of necessity, all the imagery used in this system is pre-created. Lectures are on videotape and, as Stephen C. Ehrmann has written, "Video lectures are not the same." When done semi-professionally, a taped lecture can be better than the original. Errors, repeats, fumbles, etc., can be edited out. The "script" can be tightened, and other visuals can be inserted to break up the monotony. But when done badly, where all you see is a single 50 minute, dimly lit, long shot of a professor talking, the results are fatal. 

Other considerations are "political" and go to the heart of campus-bound direct instruction. In Maryland, where the Commission on Higher Education is considering allowing students to receive their degrees entirely through distance learning courses, half of the proposed courses will be time-independent and taken using interactive video. Not everyone in Maryland is happy about this prospect. Although distance learning will enable the state's widely distributed students to get a college education at a fraction of the cost of dormitory, campus-bound students, some higher-education leaders worry about the quality of these technological "solutions." Others worry that all the politicians see is dollars. Time-independent distance learning saves dollars, and, once created, could eliminate the need for many professors and expensive campus facilities. 

Political considerations to one side, time-independent distance learning does not have to be all bad. Access to the needed material may be restricted to a point source such as a classroom staffed by an adjunct which is open either for a limited time or on a "drop in" basis. Libraries on campus may provide students with access to all or some of the materials either all the time or during special hours. Some of the materials--VCR tapes, books, computer programs, simulations--could be placed "on loan" in a library and only bona fide registered students allowed to borrow them and take them home for further study. By controlling the access, educators can remain in control of their material and how it is used. 

In some respects, time-independent modes of distance learning favor non-linear modes of education. Although traditionally linear classes and subjects can be provided in a time-independent manner, non-linear, interactive classes and subjects have a distinct advantage. Once the sequence of topics within a course is placed in the hands of the student, the syllabus must consist of complementary topics and ideas, each of which strengthens the whole. Sometimes, when forced to think this way, educators devise better, more rigorous, but more accessible courseware and the quality of education rises. 

Time-independent modes of distance learning also promote a high level of interaction between the students and the subject. Study after study has shown that such interaction favors increased quality and quantity of learning. A typical module might consist of a video "short," an interactive computer program that allows the student to discover the concepts from a variety of approaches, an analysis or exercise segment to test the comprehension so far, and the ability to e-mail (or equivalent) the instructor for follow up or testing purposes. A whole course would consist of 10 to 14 such modules and be accompanied by textbooks, workbooks, and a full range of examinations. 

Delivery Mechanisms 

Broadcast: Milton Keynes is a university town in England that services thousands of students a semester. A visitor to this modern, grass-covered town, however, might be surprised at the total absence of undergraduates. This is because Milton Keynes is the site of one of the most famous distance learning projects in the world: the Open University. From silent halls the Open University sends out its lectures by means of a standard VHF television broadcast signal. Anyone with an antenna and a TV set can get at least part of an education. A pioneer in many ways, the Open University uses the oldest and most general means of reaching its audience: broadcast television or, these days, satellite relay. Both of these are effective delivery mechanisms with advantages and disadvantages for distance learning. 

Delivery mechanisms fall into two broad categories: broadcast and distributed. Almost all distance learning projects use a mixture of both. As the name suggests, a broadcast system develops and places a signal onto a broadcast medium such as electromagnetic radiation (the VHF TV signal), or amplitude waves down copper or fiberglass cable. At one time, the TV broadcast companies could not control who received their signals so television was "free" to the recipient. Today, with various encoding mechanisms, the signal can be scrambled so that, in theory, only the legal recipients can decode and enjoy the product. The same is true for satellite broadcast or cable broadcast systems; the producer uses a "general" delivery mechanism but tries to limit the receivers to legally entitled consumers.

Obtaining an FCC license for broadcasting a signal over the general frequencies used by TV companies (VHF or UHF) is not easy and is quite expensive. Usually any distance learning project which plans to use such a mechanism would either team up with local public broadcasting stations (as they do in England and parts of the United States), a distribution company (like the BBC or Fox broadcasting companies), or have a powerful organization such as the state or local government handle the broadcasting of their signal. 

Satellite transmission and reception are easier to arrange, but even here the transmitting participant must have a full-time broadcast engineer and be prepared to invest heavily in expensive front-end equipment and service personnel. Laying cable is a convenient way of delivering a signal at the local level, but becomes expensive as the distance increases. Few institutions try it outside of the commercial cable companies. However, these days most potential recipients have low bandwidth telephone cables within easy reach, and many now have high bandwidth cable TV outlets in their homes and offices. Distance learning providers, therefore, often form partnerships such as the one between Penn State and AT&T, in which the provider (the university) develops the signal and the company (phone company or cable operator) delivers the signal to its customers. 

A broadcast signal, either from a transmitter or through a cable, used to be a time-dependent method of delivering its messages. Open University students were once famous for getting up at 5:00 in the morning to get their next lecture on "Modern Poets and Their Influence on Einstein's Theory of Relativity." VCRs have changed all that. Students can now program their machines to "download" a lecture from a commercial broadcast medium at the time of delivery (which for obvious reasons will not be "prime" time!), and study its message at their leisure. 

Distributed: The New York Times, Microsoft, MGM, and W.C. Brown Publishers also need to deliver their products to the consumer. These companies use the time honored method of pattern distribution to accomplish this feat. With many variations, pattern distribution systems work on the general principle of selective matching. Microsoft creates Windows 95 software and distributes its disks in computer stores (this is the "pattern," you do not place your computer disks in shoe stores). Recipients with DOS based computers are a natural "match" for this software and purchase it. It would be a mistake, and the system would fail, if Macintosh computer owners were to be given Windows 95 (they have been using a better product for years). All commercial distribution methods work roughly the same way: pattern distribution and selective matching.

Distance learning providers must also use distributed modes of delivery for at least part of their product. Text, certain images, sounds, software, VCR tapes, animation, tests, answers, etc., all or in part must be distributed selectively to the appropriate students. If all the students can visit the campus at one time or another, the pattern distributor could be the college bookstore. If the geographic area covered is larger, the U.S. mail or local outlets could be the pattern distributor. But educators should be aware that "mail order" services are notoriously difficult to provide and to run in a cost efficient and trouble-free manner. As with the broadcast situation, some universities turn the "mail order" aspect of pattern distribution over to professionals. 

Additional complications arise when issues of timing and security are involved. If it is possible, all materials such as workbooks, videotapes, software, etc. , can be pre-distributed before a linear, time-dependent course begins. But when courses become non-linear, and/or time-independent, the distribution pattern may become more and more chaotic as students demand partial distribution at almost any point in the total presentation. This problem is especially acute when testing is involved. Examinations and the results of these examinations need a high degree of selective matching during distribution. Even within a closed campus, issues of examination security are a constant problem. With distance learning, not only must the right (and legal) student be properly matched, but the security of the examination process must be even more vigilantly guarded. At the Open University, every year students would get on trains all around the country and travel to designated sites to take final exams (the system of testing is very different in English universities). 

It is not clear that present distance learning projects have fully resolved all these issues. In part, this is because distance learning is such a new concept that not all the finer points have been worked out, and not enough faculty are yet engaged. As, or if, this mode of education grows, formulas will have to be developed to assist faculty with the larger problems of delivery. At the moment, ad hoc schemes of delivery cause more frustration among distance learning faculty than any other issue. During distance learning, many aspects of teaching protocols are taken out of the hands of the teacher, but none more so than the technologies of delivery. Disillusionment builds with this loss of control. Universities, and other institutions starting to consider distance learning, need to address the "delivery" problem early in their deliberations and ensure that the whole system does not fail because no one planned this critical phase correctly. 

Internet: It is almost impossible these days to talk about delivery mechanisms without someone mentioning the Internet. A full discussion of the past, present, and future of the Internet is beyond the scope of this paper, but a couple of points should be raised. Despite all the hype and all the excitement, the Internet is basically a very narrow bandwidth, general broadcast system that was originally used by academics and others to pass text messages in a linear manner. With clever improvements in data transmission, compression, and search engines, it is now possible to retrieve from "the net" text, sound, still images, and moving images, some of them even in "slow" real time, but the basic limitation still remains. The Internet is a broadcast telephone system on steroids. 

No doubt this will change, but at the moment try logging onto a site any time after 10:00 in the morning and the second limitation becomes obvious: overload. As commercial providers have made "the net" available to everyone who can pay their fees, limited resources have been strained almost to the breaking point. Since the substructure of hardware, software, and personnel to run the Internet is almost voluntary, this overload is rapidly approaching crisis proportions. 

Perhaps this will cause a collapse or a total revision of the Internet, but it would be a mistake for any distance learner to depend on the current Internet to remain the same for much longer. 

With these limitations in mind, it is still possible to take advantage of the unique features of the Internet, particularly as expressed by e-mail and the World Wide Web. E-mail is rapidly becoming the communication medium of choice for a wide range of users. Among these are distance learners. Rapid point-to-point delivery of pattern distribution with selective reception are both possible at one and the same time on the Internet. This feature alone makes e-mail and the Internet valuable players in distance learning. A professor could pattern broadcast a message to 20 students or 2000 students on e-mail and receive 20 to 2000 point-to-point individual replies, all without leaving the keyboard. 

The World Wide Web is an environment that can perform certain functions better than any other medium. "Home pages" on the Web are a unique form of general broadcast that allows almost anyone with a minimum experience, a word processor, a Web browser and an Internet provider the ability to "broadcast" his or her message. You don't need a license, broadcast engineers, fancy software programmers, consultants, or designers. All you need is something to say and a little time to say it. 

For distance learning this kind of "broadcast" has obvious uses. A home page can become a "virtual university" where potential students can discover what programs are offered, what courses are being taught that semester, what the syllabus is, where to register, what forms they need, and what rules they will have to follow. Many campus-based universities are now turning over almost all their information distribution systems to the Internet with vast savings in time, money and paper. 

Each "course" given by distance learning can have its own home page where students can selectively find course information, text, images, software, and even pre-tests for their subject. In many ways, having a home page for a course gets around some of the pattern distribution problems mentioned above. But it is not a total solution. Commercial publishers object most strongly and properly to professors "posting" whole sections of their textbooks on the Web. Copyright is a slippery concept these days, but freely distributing a work of literature, art, or music over the Internet in violation of the rights of the author is not only unethical, it is probably illegal. However, as more and more customized materials become available from professors and commercial suppliers, this problem may resolve itself. 

Return to Table of Contents

Since graduating, Nick has instructed adult students in computer techniques and operating systems at the Consortium for Worker Education, managed data processing for the Adult Literacy Initiative Evaluation System, taught as an adjunct lecturer at our own City College providing instruction in basic microcomputer applications, and, most recently, provided corporate training in systems, networks, the Internet/WWW and desktop publishing in and around the NYC metropolitan region. 

As facilities manager, Nick will be responsible for supervising and providing leadership for the staff supporting the three Academic Computing facilities: the Faculty Training and Development Laboratory; the two new multimedia classrooms located next to Lab in the basement of the Library (their grand opening took place at the beginning of the Spring 1997 semester); and the soon to be completed 24 hour electronic study area, the Library Café. 

In addition to these day-to-day responsibilities, Nick will also be planning a comprehensive program of faculty development workshops, assisting faculty in developing distance learning modules and courses, teaching faculty a wide range of electronic applications (through personal appointments in department offices or formal instruction in the Lab or classrooms), and, in conjunction with the Library's instructional program, teaching and training students in the use of electronic resources. One can only assume that Nick will be very busy, indeed. 

Clearly, Academic Computing has found a treasure. Please feel free to visit the Faculty Lab to talk to Nick at any time. Or, if you prefer, call (x4634) or e-mail (niron@brooklyn.cuny.edu) him for advise and/or instruction in any computer applications that are of interest to you. All of us in the Library wish Nick the best of luck and many years of professional satisfaction here at Brooklyn College. 

Some speak of distance learning, others flexible learning, still others time-independent learning. These terms (and other related ones) apply to educational experiences delivered to students located at a distance from the instructor. The technologies and settings will vary. In some cases, classes are delivered via telecommunications--the Internet or the World Wide Web. Students may take the courses from their homes using personal computers, or travel to a specific site (a local high school, perhaps) where they will find the equipment and other support services they need. Teleconferencing (satellite, cable) is another approach to distance learning. Simpler, very basic methods can include mailing videotapes back and forth, and telephone or audio-conferencing. 

Technology: The Vehicle for Distance Learning

Whatever the term one chooses, experts caution that institutions venturing into distance learning avoid overemphasizing its technical aspects: technology is merely the vehicle for distance education: it is temporary, and it is certain to change. Rather than investing a great deal of time struggling with distance learning's technology, when local expertise isn't there, a consultant can get the project started. This allows the institution to move forward with course design and other issues that require local intellectual capital. 

Many schools have elected to use "lowest common denominator" technology (286s versus 486s, for example), in order to teach as many students as possible. It is important that both materials and delivery systems are accessible and easy to use. 

Why Distance Learning? The Student Perspective

Nationwide, 20% of today's student body is under age 20. Older students are often busy with jobs and families. In rural areas, distance learning's clear appeal is that it provides educational opportunities where none existed before. In crowded, urban settings, students may live only a few miles from campus, but traffic and parking problems can make their commutes very unappealing. 

Convenience has become a major factor when students make decisions about where to study. Those for whom the cachet of a prestigious school is important may continue to choose on-site, traditional education; they will be willing to pay more to take a course at a well-known university. Others, who value accessibility and reasonable cost, will find distance learning attractive. There are signs that some students are beginning to think about schools that offer the option of "one stop shopping" they take their classes from home via computer. Afterwards, from a single 800 number, they get the library, career counseling, student aid--everything--that supports their educational experience. 

Distance Education's Appeal for Colleges and Universities

Modern technologies have erased the traditional geographic boundaries that defined many colleges. Once a school enters the sphere of distance learning, its courses can be offered to traditional populations, as well as to students living in other cities, states, or countries. Thus, distance education offers institutions an opportunity to expand their enrollments.

Competing for Students in the Distance Learning Environment

In a competitive learning environment, it is important that schools be flexible, positioning themselves so that they are capable of serving students in many ways and at any time. Some higher education critics suggest that colleges that want to survive and prosper in the new century must become much more customer-oriented. 

Institutions must understand that, whatever their individual decisions about distance learning may be, the very existence of this new educational model will cause the "place allegiance" their students and faculty once felt to diminish: over time, each group will begin to see itself as independent contractors, free to take courses and offer courses anywhere. Schools with distance offerings will begin to siphon students from traditional colleges. (In some states, where there are more students in the pipeline than there are buildings and faculty to educate them, this is good news. In others, where competition among institutions is already acute, it is not.) 

There are other institutional rewards for schools that choose distance education. When faculty work together to plan a full distance learning program--a group of linked courses so that the school can offer a degree via distance education--benefits also accrue to the traditional curriculum. Many schools require that everyone hired teach in both the traditional and distance learning programs. 

Which Courses Are Most Promising for Distance Education?

It's important for a school to position itself strategically for distance learning to determine its competitive edge, and take advantage of it. Schools might think of distance learning offerings as a product. As one specialist in the field suggests, Pepsi and Coke compete because cola is cola: before long, courses will be courses. This means that each school will need an image, something to distinguish its offerings, to make them attractive in the market, to cause students to think, "These are nice people and I should enroll with them." The message must be, "The student comes first." 

There is no course that cannot be taught via distance learning--even public speaking is a possibility, using voice mail, videotaping, and mailing tapes back and forth. Most experts advise that institutions mainstream their distance learning courses to make them a part of the regular curriculum, rather than treating them as a separate program. The idea here is that every student, at some point in her or his career, will find it inconvenient to come to campus and want to take a course via distance education. 

There are two approaches to determining what one's distance learning course content will be. In the first, the school assesses what courses its students most want: business? nursing? the required or "core" curriculum? (A variation on this strategy calls for collecting data about the students the school wants to attract, the courses and support services they require, then proceeding along the lines indicated.) Using the second tactic, an institution proceeds in the areas where its faculty strengths lie, and offers courses in whatever they are teaching. 

Whatever courses a school elects to offer via distance education, advertising and marketing the classes--deciding what the message will be, how and where it will be communicated--is critical. Online marketing can be done via a school Web site. Faculty can also announce their courses through commercial providers. 

Faculty Training for Distance Learning

It is not uncommon to spend 50% of the budget for distance education on faculty training. Initially, faculty may feel uncomfortable with the idea of distance learning. To ease any awkwardness, one model for faculty development has the instructor offer the course in the classroom one semester, then via distance learning the next. Faculty must also be trained in adapting traditional courses for the distance environment: developing these courses takes more time than designing traditional ones, and requires new skills. Faculty will need to learn ways to ensure the quality of class contributions in the remote setting; they must develop techniques for sequencing assignments and structuring distance discussion. Clearly, a system of faculty training is an important underpinning for a distance learning curriculum. 

When video is involved, some schools provide faculty with training that will make them better presenters. This may involve coaching in appropriate dress and hints about speech. Today's students are so accustomed to high-quality commercial television that they expect the same excellence in video distance learning. They may suffer through mediocre presentations, but if they have a choice they will choose something better. 

Intellectual property is another important issue when faculty develop distance education classes. Who "owns" the distance learning courses that are developed at an institution? Two patterns predominate. At some schools, policy provides that everything developed at the institution belongs to the institution, because its employees produced it. At other universities, private ownership is the accepted model: the person who develops the course owns it and can take it anywhere. In the first case, the institution is responsible for any legal problems that may develop. Once a course is designed, faculty other than the developer may "teach" the course. The developer may be paid to update and revise his or her courses on a regularly scheduled basis, perhaps every three years or so. 

Faculty Rewards: Why Would Anyone Want To Do This, Anyway?

A common refrain when faculty are asked about developing distance education courses is, "I'd like to do it, but I won't get my promotion unless I get this paper done." New faculty rewards are needed to create the time to develop distance learning courses. A school might consider some of these incentives: 

• Faculty can be given release time for one course, the semester before they are to teach a new distance learning class, giving them time for course development. 
• As an alternative, distance learning courses can be taught as add-ons or overloads, for which faculty receive extra pay. (The danger here is that the extra work may tire faculty, who may not produce their best.)
• Institutions may pay faculty at a higher rate for teaching a distance learning course than for teaching one on-site. 
• Schools may pay faculty to adapt the courses they teach in the classroom (probably high enrollment courses) to the distance learning environment.
• Adjuncts, who may be more flexible than full-time faculty, can be hired for the distance learning curriculum. (Campuses that go heavily into distance education typically hire many more part-time faculty. Often they pay more for distance learning courses, establishing a new and separate pay scale for them.) 
• The school can offer special perks to faculty teaching distance learning courses--provide them with laptop or other computers, for example. Perks may attract the "explorers" among the faculty, the initiators, those who are willing to take the lead. (Experience shows that, once the best and most admired faculty buy into distance education, others will follow.) 

Curricular Issues

Some distance learning programs fail because the institution neglects to prepare the entire faculty, the administration, the board of trustees--everyone who has a stake in curricular issues--for this new approach to teaching. When a distance education program is developed, everyone with a rôle in curriculum must be involved from the start. Institutions should be careful that they establish the same provisions for course approval, academic oversight, and review of distance learning offerings that they employ for on-site classes. 

Diversity in content is more important in distance than in traditional education: when a school launches a distance learning program, its student body will change. It may attract students from other cities, states, and even other countries. Thus, faculty must create a curriculum that has the capacity to reach anyone, anywhere--the courses must be global in theme and content, rather than "American." It is also worth remembering that, no matter what courses a school offers, it is almost certain that someone who is not a bona fide student will manage to get in. Intentionally or unintentionally, the school's audience will be diverse, and its courses more public. For political and legal reasons, it is important to consider course content carefully. 

Faculty must also think of ways to individualize distance learning courses, despite the fact that there may be 60, 80, or more students participating: how can they make the student feel that his or her experience is in some way a personal one? Dividing larger classes into smaller discussion groups is one approach. Students in each group might receive tailored reading lists, talk online or via other technologies about what they are reading, then switch into other groups, as the semester progresses. 

How Should Distance Courses Be Priced?

This is probably a bigger issue for private schools than for public ones. The simplest approach to determining tuition for distance classes is to pick a competitive price point for the course, then develop a product to match it. Consider that, if a student can get a required subject (English 1, perhaps) via distance learning from an accredited institution, he or she may be reluctant to pay more for the home campus's equivalent. 

Course and Teacher Evaluations

Appropriate criteria for evaluating distance teaching (log on frequently; develop a distance teaching style) must be developed and accepted before faculty evaluations are conducted. Adding these later, as an afterthought, will discourage faculty participation in the program. 

Many factors will affect the ratings students assign to distance education teachers and courses. Learning outcomes in distance courses are often the same, or better, than those for on-site students. Students who have never taken on-campus courses--whose first experience with college is via distance learning--express greater satisfaction with distance education: they have never enjoyed the social interaction an on-site course provides. The more remotely located the students are, the greater their satisfaction is likely to be. Highly-motivated students with an affinity for technology generally do best in distance classes. 

Typically, students' initial expectations for distance and on-site courses are the same; in the end, however, satisfaction is sometimes lower for remotely offered classes. This may be because students are not fully comfortable with the technology of the distance learning environment. (Schools often discover they need staff to help students with connecting to and working in the online or other technical environment.) 

On the other hand, at one large university students rated distance learning classes enrolling 90 students higher than traditional courses with the same number. This is not surprising when one considers that, in the distance version of this course, students went to nearby community colleges and took classes in small groups of about 10 persons each. Within the group, students formed strong interpersonal relationships--much stronger than might occur in an on-site class enrolling 90 persons. However, the distance learning students rated the teacher (vs. the course) lower than in similar traditional courses, perhaps because they had less personal identification with him. 

Faculty who teach courses they designed themselves often have very positive distance learning experiences, those who offer courses designed by others less so. While some schools purchase "canned" courses from other sources, many institutions have learned that it is important to have faculty "own" the courses they teach. 

Some schools observe that faculty receive higher student ratings when teaching a traditional course, after they have taught the distance version. Why? Faculty who teach distance learning courses have to be well organized and present all the material in the course. Unlike on-site teachers, they cannot reach the last class and apologize for material they hoped to cover but never got around to. 

Meeting the Requirements of State and Other Regulatory Bodies

Each state has different laws governing higher education degree requirements. There are also state and regional accrediting bodies. Some educators who are experienced in the field of distance learning worry less about faculty acceptance than about that of people in these policy-making positions.

State or other review boards have barely begun to think about what it means to have a quality distance learning program--what the standards and principles will be. Fortunately, a consortium of western schools has developed a list of commonly accepted principles for distance education quality and good practice which it has disseminated to every state and regional accrediting body in the nation. The hope is that these principles wil be adopted. 

Absent special regulatory guidelines for distance education, a school must demonstrate that its offerings meet existing state and other guidelines for contact hours, attendance, and other degree requirements. 

Support Services for Distance Learners

For the student who takes the occasional course via distance learning, a full array of remotely available support services is probably not an issue: the student is coming to campus for some of her or his classes and can take advantage of those services on-site. 

Nonetheless, disclaimers may be necessary for certain distance classes-- "No tutoring," for example. And, schools may wish to offer students taking even one distance learning course options like electronic registration and online course counseling. They may also choose to develop other resources, like an electronic "reserve shelf," a librarian-on-call, Internet access to librarians, or a connection to the bookstore that permits students to order online with a credit card. 

The school that elects to offer degrees via distance education faces a completely different set of issues: it must ensure that students who study remotely have access to the same services that on-campus students receive. Without lowering its standards, the institution must adapt its procedures to the online, telephone, and fax environments. Some institutions have taken steps to ensure that it will never be necessary for a student to set foot on campus to earn a degree. In the process of developing the program, they involved key personnel in admissions, advisement, the bursar's office, and student aid. 

Assessment and Distance Learning

Evaluation and assessment must be integrated into the distance learning program. It will also be important to establish benchmarks for students, since there will be nothing like "Coming to class twice a week at 8:00 AM" to fall back on. 

Institutions might use focus groups not just to plan their curriculum and how to offer it, but also to evaluate how they are doing once the program is underway. Measuring progress toward goals is critical, and assessment procedures should be in place from the start--before the first course is taught. 

One school used questionnaires at the beginning of the semester in which it sought information about attitudes toward technology and distance learning, reasons for taking the course, goals for the class, expectations about behavior during the course, access to technology, and expertise and comfort in using technology. At the end of the term, it used a follow-up questionnaire, as well as selected telephone and in-person interviews. 

The assessments process teaches many schools that they need to spend more money on faculty development. Often institutions find that many students had no previous computer experience and took the course as much to learn about the technologies as for course content. Some students spend so much time on the technology they do not get much out of the course. In almost all cases, problems associated with students' lack of technical proficiency are underestimated by both institutions and students alike. 

A number of foundations and corporations provide support for various aspects of distance learning: planning grants, support for faculty training, funding for equipment. A school must start by defining its curriculum content. It might then draft a two-to-three page letter of inquiry, outlining its project goals, which can be sent to many places; granting agencies often find collaborative projects appealing. 

At some point, the school must develop the technical infrastructure, design faculty training, and market the program. An institution might want to start small, with three to five courses taught in the classroom the first semester, then via distance learning the second. Some schools have given a tuition break to students who enrolled in their initial distance learning courses. 

Keeping Current With Distance Learning

Until recently, information about distance learning was hard to find. Today, it's overwhelming. An institution thinking seriously about distance learning should consider creating a new position, resource librarian and analyst. 

The school will need someone to collect the vast amounts of distance education material, digest it, and tell the program's manager what's happening in the field and what the school should be doing. 

Distance Learning: Yes or No?

Some academic leaders see distance education as basic to the survival of higher education. Despite the many complexities associated with venturing into this sphere, the benefits for both students and institutions are considerable.

Distance education is a fundable idea (in terms of foundation and corporate support), projects institutional growth, and increases revenues; at the same time, it raises a school's leadership profile. For all these reasons, it is an educational concept most schools will find difficult to ignore.

Return to Table of Contents

Shopping for a PC: A Three-Step Process

The process of helping someone find a good PC is threefold. First, you send them to buy the most current copy of Computer Shopper (PC World, PC Magazine, etc.). Next, you recommend a setup you feel you yourself would buy, given the dollar constraints. Finally, you set aside two hours to justify the cost of each piece of hardware that's provided in the package, with its stand-alone street price. 

Before long, you find yourself saying things like, "That modem is good, it is top ranked in PC World," and "That monitor is great, it came out in the top three of monitors in the PC Magazine review." From my experience, you never hear yourself recommending hardware (or software) on the premise that it "never came in last when tested." You see, when it comes to spending money on a computer (whether for hardware or software) people want the best, but more importantly the fastest. 

Finding the Best Operating System for Multimedia

When I first set out to research what I would label "The Best Operating System for Multimedia," I expected to find numerous reviews on OS/2, Windows 95/NT, and Macintosh. My expectations were more than satisfied. The Internet provided a means to search most major PC magazines and supplied hundreds of articles (past and present) pertaining to this topic. The disappointment came when I reviewed each article for relevant information. I was surprised to find that, with all this information at hand, not one article pointed its finger to the "Operating System of Choice." 

This indecisiveness was nerve-racking at first, but then I came to the realization that the question was not which OS held first place, but which performed (and in some cases outperformed) consistently across all areas. The answer to this question was simple. With all the comparisons I read and reread, never did I see Windows 95 in last place. Sure, it took second in certain areas, but when all was said and done it performed best all around. With this in mind, I will show you why I label Windows 95 as my "Operating System of Choice for Multimedia."

Windows 95: My First Choice

When one thinks of the word "multimedia," audio, video, and graphics always seem to come to mind. To some, multimedia is a well designed Graphical User Interface (GUI). To others it is the ability to watch a music video on your PC or simply play a game like Myst. One definition of multimedia, offered by Fred T. Hofstetter, is "the use of a computer to present and combine text, graphics, audio and video with links and tools that let the user navigate, interact, create and communicate."¹ With this definition in mind, we can see why it is safe to say that multimedia influences all applications on the desktop today. 

In my opinion, we need to give credit where credit is due. Sure, most applications today sport a nice front end with pretty sounds, fancy colors, and ease of use, but behind all these applications lies a sophisticated environment handling duties ranging from traffic cop to filing clerk. It is the operating system and its relationship with hardware which truly makes possible multimedia in its entirety. 

Storage: The First Feature to Consider

We begin our road to Windows 95 by taking a closer look at some key features which make this Operating System (OS) tick. Our first step to understanding Windows 95 is storage. Storage plays a vital role in the success of multimedia in more than one way. The hard disk is responsible not only for storage but for memory as well. "Hard disk buffer size certainly is not very important under Windows 95."² On the other hand, this OS is extremely sensitive to hard disk speed. 

Higher disk speeds translate to better performance. That "Win95 more effectively pages portions of itself and applications to its hard disk swap file when memory runs low,"³ contributes to its success. Another advantage is that "Win95 eliminates much of the frustration--if also much of the fun--of tuning your system's performance by handling most of the hard disk challenges for you."⁴ It automatically configures virtual memory settings and intelligently manages the size of your swap file. You can adjust virtual memory settings if you want to. 

As for the size of the swap file, Windows 95 expands and contracts it as needed. Behind the scenes, we see that the swap file can use "noncontiguous storage on your hard disk and can even reside on a compressed drive managed with Windows95's DRVSPACE.VXD driver."⁵ Windows 95 utilizes VCACHE (a 32 bit protected mode cache driver that replaces Smart Drive) to manage disk-cache size. VCACHE doesn't waste time buffering swap file reads and writes. Also, it counteracts excessive paging by decreasing the size of the swap file. Overall,"Windows 95 makes it easier to get the most out of your hard drive whether you use SCSI or IDE."⁶

A few modifications in Windows 95 make the challenge of deciding which type of hard disk to use a bit harder. Windows 95's multithreaded design plays to the strengths of the SCSI drive which, unlike EIDE, handles multiple simultaneous requests smoothly. Secondly, by providing support for 32-bit protected mode SCSI device drivers, Windows 95 provides SCSI with the integrated support IDE has long enjoyed. 

Memory: The Next Thing to Think About

Memory is the next stop on the road to Windows 95. "Memory issues have always played a key role in determining overall performance of an operating system."⁷ Windows 95 does not change this basic fact. If we focus on how much memory Windows 95 uses, we quickly come to the sad conclusion that it is a memory hog. However, you can rest assured that Windows 95 uses memory in a much smarter way. 

Like its predecessor (Win3.1), Win95 still loads only portions of applications currently in use, but it "does a more effective job of swapping code to and from the hard disk."⁸ Windows 95 fools each application into thinking it is the only one currently running and utilizes the clipboard, DDE, and OLE to exchange information.

Windows 95 also improves management of memory in other ways. Heaps use larger linear address space. With the 64k (Win3.1) heap limit removed in Windows 95, you will see fewer "Out of Error"⁹ messages. Under Windows 95 you will notice that a number of data structures have been moved into 32-bit heaps. This makes it very unlikely that you will run out of system resources. 

Preemptive multitasking is another inviting feature of Windows 95. It offers a smoother, more robust multitasking environment. The engine behind this new addition is known as the thread scheduler. It is a piece of the Virtual Memory Management System which "distributes system time based on each thread's priority level and readiness to run."¹⁰

The only flaw in how Windows handles memory is that it has gone the route of backward compatibility. The possibilities of a system crash linger on because of the location of 16-bit applications. These are cast into a location of memory all their own where they are free to corrupt each other and threaten overall system stability. As we slowly see 16- bit applications disappear from the desktop, you will notice a corresponding reduction in crashes. 

Communications: What Should We Expect From an Operating System?

In continuing our effort to see Windows 95 as the choice OS for multimedia, we must stop to ask and answer a simple question: "Can an operating system facilitate communications?" The answer to this question is obvious: "Yes, it must provide the means to talk to various often used peripherals (i.e., printers)." 

Communications in the 90's means far more than simple serial data transfer between two devices--an evolution Windows 95 reflects.The new operating system has been designed to provide services for users who may need to make simultaneous connections to a variety of electronic mail, fax, or on-line services and who are as likely to be mobile as they are to be tied to a home or office chair.¹¹

In the areas of high speed reliability, data throughput, and device and hardware support, Windows 95 demonstrates its flexibility in communications. With less critical code in its kernel, communication processes do not find themselves waiting for each other to finish accessing this code. Windows 95 is better capable of responding to interruptions at communication ports as well. This is attributed to its communications subsystem consisting of the new "multithreaded, multitasking, 32-bit communications device driver."¹² Windows 95 also provides support for the use of parallel-port modems for faster communications. 

Windows 95 is plug and play compliant. Plug and play (PnP) technology "permeates everything the operating system does."¹³ One of the fundamental advances in Windows 95 that makes PnP possible is that it dynamically loads device drivers (and can do so at anytime). Windows easily detects the presence of a new hardware device, installs the appropriate driver and activates it. It even improves the process for devices like PCI adapters, which force you to power down before installing them. Non-PnP devices are not left in the dark in this enhancement. The ability of Win95 to give you a reliable inventory of IRQ's, DMA's upper memory ranges, and allocated I/O ports makes it much easier to configure these devices. 

In the area of networking (one of the most important aspects of communications), Windows 95 is far from runner-up. It has better support for PC Card network adapters. It contains a built-in Novell-Netware client as well as support for third party "real mode" and "protected mode clients." "Direct Cable Connect (DCC) is one of the greatest untold stories of Windows95. Network users with notebook and desktop computers get free and easy networking for their notebook by simply connecting it to their desktop client with a parallel cable." Windows 95 also possesses a dial-in networking client and can be installed remotely on a client.

Gamesmanship: Another Feature That Sets Windows 95 Apart

At this point in our trip, we have arrived at the topic of gaming. Windows 95 stands apart from all other multimedia-capable operating systems in its ability to satisfy power-hungry "gamers." Nothing tests an operating system's ability to handle sound, video, and animation at such high speeds as running today's games does. 

Developers of Windows 95 realized early on that to ensure its success as a gaming platform they would have to provide channels for applications to talk directly to hardware. The Windows 95 Game Software Developers Kit provided developers with the means to do just that. It is "a complete set of application programming interfaces (APIs) for creating fast-paced, graphically intensive games previously seen in DOS and console gaming worlds."¹⁴ The APIs consist of: 

DirectDraw: which allows ease of communication between developers and display hardware for operations like color-space conversion and color-keying; 

DirectSound: which allows "high fidelity, low latency"¹⁵ sound and mixing of multiple audio streams; 

DirectPlay: which enables "simple multiplayer game connectivity";¹⁶

DirectInput: which provides support for future direct access to analog and digital joystick hardware. 

The advantages of Direct-X don't stop here. Users will notice a decline in game tech support prices owing to the overall ease of use when installing, configuring, and running a game on Windows 95.

Windows 95: You Can't Argue With Reviewers!

A major factor in my decision to vote Windows 95 the best multimedia OS came from reviews that described how it compared to other operating systems currently in use today. As you will see on the last stop of this tour, Windows 95 does a great job of picking up the slack when other operating systems fall short. 

Windows NT Workstation is a high-end personal operating system built for client-server computing. With built-in security (approved by the government), controlled access at the desktop level, and a secure file system, NT is a pure Power PC. However, it is far from an acceptable operating system for multimedia computing today. NT's architecture makes it merely usable at the point where Windows 95's performance peaks. NT devours RAM and in the past has had significant problems with disk size. It is the choice for users who want the greatest amount of stability and protection. Though "Microsoft NT supports the fastest microprocessor ever and has the capabilities to access power Windows 95 can't reach,"¹⁷ when it comes to games and video, NT just will not perform. 

Out of all operating systems, OS/2 found itself at the bottom of the ratings chart time and time again. It seems that OS/2 provided the biggest challenge when it came to completing everyday tasks, and it appeared to crash the most when tested against Win95 and NT. Another flaw in OS/2 is the fact that it uses a "single queue to manage all input from the mouse and keyboard. If an application fails to retrieve input meant for it from the queue, all other running applications are forced to wait, which can freeze the O/S."¹⁸ As with the Macintosh, OS/2 performed significantly more slowly when testing office applications like MS Excel. It suffers from a limited supply of drivers and lacks up-to-date applications (6400 as compared to Windows 11,000). 

Though steadily losing its market share, Macintosh is still known as the choice PC for user-friendliness. Macintosh provided the drawing board from which the Windows 95 desktop borrowed most of its design ideas. As far as I am concerned, Mac still holds the market in desktop publishing and animation. Based on the RISC architecture, Macintosh is a popular OS for cross platform graphics. However, in the area of business-computing it leaves a lot to be desired and performs sluggishly within Excel and Word. 

Windows 95: What Will the Future Bring?

Though our tour into Windows 95 is near its end, this article could not be complete without a look at the future of Windows 95.Future versions of this powerful OS will see support for new forms of high-speed communication, as well as support for the IEEE=1494 Bus Interface Standard.¹⁹ Apart from little GUI improvements, Nashville (the future version of Win95) will have only a Personal Information Manager (PIM) and a small number of device drivers added to it. 

When considering the best PC for multimedia, I decided to stray from the norm and pursue the operating system which wasn't always tops but at the same time did not live at the bottom of the ratings. My arguments for Windows 95 as the best operating system for multimedia computing are plain and simple. 

• In today's world, speed is everything; it is the first factor we look at when answering questions like, "Which OS is the best?" 
• Windows 95 demonstrated its ability to handle multimedia as the culmination of ordered chaos. It outperformed some of the other operating systems and always showed consistency. This feature I found to be very important. It made me realize that a traffic cop on the job in a busy intersection could work wonders to clear up a jam, but if he handled traffic differently in various types of weather, he could not be relied on. 
• Windows 95 is a robust operating system capable of handling all the requirements of a sophisticated contemporary computer game. 
• Windows 95 does not stop at the present--Microsoft is moving forward, enhancing Win95 to ensure that it continues to maintain its seamless environment in the midst of a fast-paced, ever-changing world of technology. 

Why would you even think about considering another operating system over Windows 95 for multimedia purposes? Windows 95 is itself a multimedia operating system. It provides all the tools to develop and run effectively mainstream multimedia applications. 

NOTES

¹Hofstedder, Fred. Multimedia Literacy. McGraw Hill, 1995. Return to text.
²Klare, Matthew. Internet: Ziff Davis, "Storage." Return to text.
³Ibid. Return to text.
⁴Ibid. Return.to text.
⁵Ibid. Return to text.
⁶Ibid. Return to text.
⁷Smith, Gregory. Internet: Ziff Davis, "Memory." Return to text.
⁸Ibid. Return to text.
⁹In Win3.1 "Out of Memory" usually resulted from opening too many windows, thus causing Win3.1 to store excessive information pertaining to each window on the heap. Return to text.
¹⁰Smith Return to text. 
¹¹Derfler, Frank J., PC Magazine, "Communications in Windows 95," 11/07/95. Return to text.
¹²Ibid. Return to text.
¹³Allchin, Jim. Microsoft Web site, "Windows 95 Q+A." Return to text.
¹⁴Ryan, Michael, PC Magazine, "Fun and Games in Win95," 09/95. Return to text.
¹⁵Ibid. Return to text.
¹⁶Ibid. Return to text.
¹⁷Lourderback, Jim, PC Magazine, "Wrap-Up." Return to text.
¹⁸"OS/2 Warp: Another Route to PC Productivity," PC World Online, 02/96. Return to text.
¹⁹Enables high-performance multimedia connections. Return to text.

Return to Table of Contents

Since the spring semester of 1992, several programming courses at Brooklyn College have made extensive use of e-mail and network facilities in connection with instructor-student communication. This has allowed us to overcome several problems related to the external demands placed on working students, limitations on computing facilities and faculty overload. 

Although in retrospect the advantages of this arrangement seem obvious, a recently conducted informal survey of some 50 computer science faculty from as many institutions revealed that most made only incidental, if any, use of e-mail and other network facilities, even though such facilities were readily available. This is a bit odd, considering that fields outside of CS are beginning to use these facilities extensively (for a good review, see Brookshire [1991]; for a description of e-mail and conferencing facilities in a computer literacy course setting, see Berman [1992]).

One important component of our use of the network involves an automated homework program checker. There have been several reports about such checkers in the computer science education literature (Kay [1993], Burris and Darr [1988], Isaacson and Scott [1989], Reek [1989]). But many of these were not network based and more importantly they were used exclusively as a homework grading aid, not as a student learning tool as described below. 

The purpose of this paper is, therefore, to alert the computer science education community to an extremely useful instructional device as we have experienced it. 

The Context

Brooklyn College's students are an ethnically diverse group of working class students, many of whom are immigrants or members of minority groups. Many are considerably older than 22 years. A significant number (well over half) work more than 20 hours a week and/or have parental responsibilities. All are commuters: our college has no dormitories. 

The hardware facilities of the college are ample but not terribly diverse. We have a network of 35 Sun IPCs, a much larger Novell network of PS-2s and PCs, and access to an IBM 3090 mainframe. There are no experimental, super-computer or parallel machines on campus. 

The courses in question all used the Sun IPCs as their base machine. These courses are: 

• a liberal arts "core requirement" computer literacy course for nonmajors; 
• a second programming course for majors; 
• an undergraduate elective course on Unix and network programming; 
• a masters level course on parallel programming. 

Existing Problems

Several specific problems were overcome by using e-mail and other network facilities: 

The discrepancy between faculty and student schedules. Due to the students' work and family responsibilities, there is no reasonably sized set of faculty office hours that is possible (let alone convenient) for all students in any given class to make use of. 

The limits of available computing facilities. Some courses require or would benefit from either hardware that is not available on campus or expensive software of which only a single copy exists on the faculty's research network. 

The slow turn-around of homework assignments. Many classes meet only twice a week. That means that the cycle of assigning a program, receiving a hardcopy source listing and output, and returning it to a student takes a minimum of a week. This long cycle makes it difficult to give the student an opportunity to redo his or her work in time to be relevant to the current class work. 

Increased faculty load. Because of budget cuts, faculty have been compelled to undertake higher teaching loads, both in the number of courses taught and in the number of students in their classes. This change makes it more difficult for faculty to comment critically on programming homework assignments or to ask students to resubmit problematic ones. 

It is true that all of these problems have political solutions (for example, an increased budget could reduce faculty load and provide more equipment), but an individual instructor does not have the opportunity as an individual to achieve them. The solutions presented in this paper, though partial, are technologically-based and available to any interested instructor. 

Using E-mail and the Network

Starting in the spring of 1992, an increasing number of courses allowed or required students to submit programming assignments via e-mail. In my second semester programming course for majors, for example, students were required to submit all their assignments electronically and encouraged to e-mail their questions about the course (including assignments, lectures, readings) to me. I, in turn, "guaranteed" a 24 hour response time. In addition, all homework assignments, homework solutions, sample quizzes, answers to sample quizzes and in-class exams, and classroom examples were posted on the network and available to the students exclusively through that medium. 

Creating an E-mail Culture

Not all students who are new to e-mail approach it with eagerness. This was especially true of the course for non-majors, but also for some of the students in the second semester programming course. To overcome this, on the first day of class all of my students were given the assignment to submit their resumes via e-mail. By responding speedily and in a personal way to the information in these submissions, an e-mail rapport was rapidly built up between instructor and student. In many cases, students responded to my response, thereby initiating a conversation that, in many cases, lasted through the whole semester. 

The incentive to make use of this environment remained strong as it became clear very quickly that the fastest (and sometimes only) way to get information (assignments, announcements, solutions) was through the network. 

Besides rapidly creating a willingness to use e-mail and network facilities, this had an additional effect. It would not have been possible to get to know so many students so quickly in an ordinary class-room setting or office hour setting, both because of time considerations and because students, like many people, are more reticent in person than in e-mail. 

The Homework Checker

Being in e-mail contact with my class mitigated a number of the problems mentioned above, but in some sense it increased my already heavy workload. Much of my e-mail reading was devoted to partially done assignments that would fail on input cases not tested by the student. Much of my e-mail writing was answering (and trying not to be testy!) such submissions.

To overcome this, a homework checking program was set up. The student executes a script that wraps any number of files of any nature and e-mails them to a special account that uses a forward file to pipe the mail to the program checker. The checker creates a temporary testing directory in which the files it extracts from the mail are placed. Upon receipt of the student mail, the checker analyses the student's homework submission and within minutes sends a response to the student and instructor. The checker's analysis of the students' submissions is driven by a "homework configuration file" that allows the instructor to specify: 

• the particular files required in the homework submission;
• the files that should be compiled; 
• the libraries that should be created; 
• the object files and libraries that should be linked to form executables; 
• the executables that should be run (and their input and arguments); 
• the run-time resource limits (execution time, maximum output size) on the executables; 
• the files to compare with the output of the executables that were run; 
• filters to apply to the output prior to comparison (e.g. remove blanks, tabs, etc.); 
• the exit status code of particular test runs of the executables;
• any functions or other externals that the student code must or must not contain; 
• scope requirements for C modules (to enforce static storage class for non-exported functions and variables in particular modules);
• shell scripts to conduct special customized tests of the homework submission. 

If the homework submission passes all the required tests, the student receives a message that "preliminary checks have been passed" and that the homework is being forwarded to the instructor for further consideration. If any problem is detected, the checker immediately sends mail to the student indicating this. Very little detail is given as to the nature of the failure. Messages such as "cannot link", "incorrect output", "abnormal termination", "missing file: something.c" are typical. Only one failure is indicated--the checker stops as soon as a single failure is discovered. 

The instructor receives far more information than the student. In cases of success, source codes (and any text files for that matter) from the student are sent, along with "nm" and "ar" listings of any object modules and libraries. In case of failure, the instructor is sent all of the above, as well as a more detailed reason for failure and "diff" listings of the student output vs. the correct output.

Receiving the failures is pedagogically crucial because the instructor can readily see what problems the students are having and most importantly what concepts have not been successfully conveyed in class. This can be corrected in the next class or, more often, by e-mailing a broadcast to all the students in the class. 

The treat, from the instructor's point of view, however, comes with the successfully completed homework assignments, I know that they compile and link properly and execute successfully on all the test cases that I provide. If, for example, an assignment calls for them to write their own string handling routines, I know that there are now strcpy() and strcmp() library calls buried. I read the students' programs focusing on overall design, documentation and usage and I respond accordingly by mail. When I do accept a program, it is by invoking a script that automatically checks for plagiarism (see Appendix II) and updates a roster on the network that the students can monitor. 

Results

Overall and not surprisingly, the most important effect this environment had was in improved teacher-class communication. As soon as I became aware of misunderstanding concerning lecture material, an assignment or anything else, I posted it on the network. Students who might otherwise, because of their schedules or shyness, not have communicated concerns and questions were able to do so. A considerable amount of class-time that would otherwise have been devoted to handing out and going over assignments and solutions was saved. Furthermore, students who missed or were late to a class were able to get their assignments accurately with no cost to the instructor. 

The impact of the homework checker included both anticipated and unanticipated benefits. As expected, students appreciated the immediate feedback from the checker and I appreciated the ability to look at their programs from a design and documentation perspective, rather than one in which I tried to imagine whether the program worked or not. In addition, because many of the checking programs expected program output and input files to be in a very particular format, this entire arrangement forced students to pay closer attention to specifications. By the same token, I was forced to do the same--and be perhaps clearer than in the past in my assignments. Any ambiguities were rapidly discovered and corrected as a result of this arrangement. Students learned not only to follow specs, but to read them critically. 

Because of the homework checker and the generally improved environment for communication it was possible to give more programming assignments and be stricter on time-limits. Being strict on time-limits made it possible to post solutions to the problems earlier than would otherwise have been the case. This made it possible for the homework solutions to serve as a genuine supplement to class work. 

One curious side-effect of being a commuter institution is that even those students who do have the capability of being on campus every day of the week often fall into a pattern where they, like their working and parenting counterparts, come only on days when they have classes. A gratifying aspect of the network environment is that it seems to reverse this trend, that is, a surprising number of students were coming in to work on their assignments on a daily basis. 

Concerns

There are a number of concerns that are raised by this approach. The most serious of these is whether the approach is an instance of technology fetishism, the inappropriate substitution of technology for needed human action. Clearly that would be the case if this approach were used to replace direct professor-student contact. Less clear is its use in augmenting such contact in a climate of diminishing resources. For example, can solutions of this kind be used to justify further cutbacks? 

The program checker, with its five minute (or less) response time raises the spectre of students mindlessly trying one thing after another to get the program accepted, and thus reduced to a condition of, as Dijkstra put it, "pavlovian slobber." If this turned out to be a problem it would be quite easy to modify the script to count submissions and reject or penalize after a threshold. In our experience, pavlovian slobber was infrequent and where it existed, it was generally the least of the students' problems. 

Finally there are security concerns. Letting a student submit a program that is run on another account poses an obvious potential problem that should be approached with caution and the assistance of a systems administrator. 

Conclusion

In summary, by making use of e-mail and network technology, the problems described at the outset of this paper were alleviated: 

• Although many students could not see the instructor during official office hours, communication in between class was extensive; 
• Turn-around of homework assignments was much faster than in a course not employing e-mail; 
• Students were able to use computers not normally available to them, either on campus or at remote sites; 
• Faculty time spent on repetitive tasks relating to increased numbers of students was reduced. 

Availability

This environment is available from the author via e-mail request. Currently it assumes a network of workstations running SunOS 4.X or Solars 2.X. It requires the presence of a special course account that contains scripts that are automatically executed by the students' login process. Plans are underway to set up a straight-forward installation script and make this environment available via ftp and the Web. 

References

Berman, A. M. "Class Discussion by Computer: A Case Study," Proceedings of the 23rd SIGCSE Technical Symposium, Kansas City (Mar. 1992). Return to text.

Brookshire, R.G. "Electronic Bulletin Boards as Teaching Tools in a University Setting," Proceedings of the Tenth Annual Research Conference, Office Systems Research Association, Washington, DC (Mar. 1991). Return to text.

Burris, H., and M. Darr. "The PROGRAMS Package for Integrated Grading," Program in Computing, Department of Mathematics, University of California, Los Angeles (1988). Return to text.

Isaacson. P. C., and T. A. Scott. "Automating the the Execution of Student Programs," SIGCSE Bulletin vol. 21, no. 2 (June 1989). Return to text.

Kay, D.G. "Don't Give Grades Without It: A Comprehensive Automated Grading Assistant for Student Programs," SIGCSE Bulletin (1993). Return to text.

Reek, K. A. "The TRY System, or How to Avoid Testing Student Programs,"SIGCSE Bulletin vol. 21, no. 1 (Feb. 1989). Return to text.

Below is an example of a homework-checker configuration file. This was an assignment in which the student had to write two source files and use them to create a library. Along with the configuration file, the instructor had to prepare the following files: