[Top of Page]

JB: --time to discuss the leading core technologies in your future. This is Judith Boettcher, your CREN host for today, and I'm pleased to welcome Howard Strauss from Princeton, our technology anchor for Tech Talk. Howard is a well-known Web and all-around technology expert. Welcome, Howard.

HS: Thank you, Judith. And thanks again to CREN for choosing me as the technology anchor for Tech Talk. The job of the technology anchor is to engage our guest expert in a lively technical dialogue, and will answer the questions you'd like answered, and to ask those very important follow-up questions, too. You can ask our guest expert, Mark Resmer, your own questions by sending e-mail to expert@cren.net anytime during the Webcast. If we don't get to your question during the Webcast, we'll provide an answer in the Webcast archive. While you were watching desktop computers get cheaper and more powerful and making whatever you had on your desktop obsolete, you probably noticed that laptop computers were becoming nearly ubiquitous. That may be because many of them are now every bit as capable as all but the highest end desktop machines, and that the price gap between desktop and laptop computers continues to narrow. For example, I just noticed that there's a computer called the Winbook. It's a 233Mhz 1.6 gigabyte hard drive machine that sells for $999, quite an incredible laptop. But there's many other reasons for the popularity of laptops and their even smaller brethren, PDA's. The mobility and portability of these small computers has the potential to change the way computing is done on campuses, in commerce and at home. The switch to mobile computers is not just a replacement of big computers with little ones. It's the replacement of an old computing paradigm with a new one that may prove to be as revolutionary as the replacement of mainframes with servers and personal computers. Just this past Tuesday, during the planning session for this Webcast, I was in West Palm Beach. Judith was in Tallahassee, and our guest expert, Mark Resmer, was in his car with a colleague, Frank Panci, heading for the Golden Gate Bridge. The fact that we could collaborate naturally over the phone together under these circumstances probably surprises no one listening to this Webcast. But Frank, although he was a passenger, could not check his e-mail, check his Web page links or modify a Web page while he traversed the Golden Gate Bridge. We can all see how that is likely to change rapidly, just as we can see how soon every student at every seat in every classroom will soon have a wireless laptop connection. Or can we? Where is mobile computing going, anyway? Which uses of mobile computing are silly and which make sense? And how do universities get mobile computing started on their campuses and keep it going? We'll address these questions and others with Mark Resmer on this edition of Tech Talk. Judith?

JB: Well, thank you so much for sharing that experience during our preparation time, Howard. It was quite fun as we occasionally would lose Mark and Frank as they would go through the various tunnels. As you have already, let's introduce our guest expert today, who is Mark Resmer, the associate Vice President for Information Technology at Sonoma State University, and now serving as director of the EDUCAUSE NLII Instructional Management Systems--often known as IMS--Project. Mark has been active in developing strategies for instructional applications of information technology for nearly 20 years and has participated in creating some of the earliest tools for Internet navigation. Welcome, Mark, and thanks for joining us here on CREN Tech Talk.

MR: Thank you and good afternoon. I'm delighted to join you here. I'm, in fact, living the dream. Not only did we prepare this talk in the car over a cell phone, but I am currently speaking to you from an extremely virtual office, looking out over Silicon Valley, where I've just stepped out of an IEEE meeting that I'm attending. So we are, in fact, doing this entirely virtually.

JB: Well, thanks very much, Mark. Let me also remind our guests that, after the session is over, if you said something really remarkable that they'd like to go back and see, and I'm sure it will be quite a bit of things that will be remarkable, that they can pick up archive sessions at the CREN Website, and the transcripts are generally out by about Tuesday of the following week as well.

HS: Okay, Mark, why don't we begin our session on mobile computing network strategies by talking a little bit about some of your experiences at Sonoma State. When did you first require students to have 24 hour access to computing and when did you require them to actually have computers?

MR: Well, in 1995, we started what we call our Short Access Program, which was, in fact, a requirement that every student have 24 hour access to a networked personal computer, and we were one of the first two public universities in the United States to do so at that time, the other being the University of Minnesota at Crookston, which was the original Think Pad U. Our program was phased in over four years, and so this is actually the first year where the requirements applies to every single student on the campus, and we're now starting to see some interesting changes, both in the kinds of machines that people are using and in the ways that the machines are being used now that the program has, in fact, become institutionalized over a significant period of time.

HS: One thing about the 24 hour access, I mean, everybody would like 24 hour access. But that sort of implies that your network stays up 24 hours a day, but it really doesn't do that, does it?

MR: Well, yes, I'd love to tell you, but unfortunately there's this little lie detector here which I have a feeling would go off if I claimed that was the case. The reality is, though, you raise an important point, which is that things that we have historically thought of as being in need of enterprise-grade quality--for example, administrative systems--are now being joined by a number of other facilities. And some of our institutions have started to recognize this and others haven't yet. I think all of us need to recognize that today, it's probably just as important that the Web server is up at 3:00 in the morning as it is that the mainframe is running and crunching the budgets at 3:00 in the morning. And that is certainly true of networks. We have, in fact, done a great deal of work on network reliability so that, while we don't have 100% up-time, it is largely the case that the network is something that people can, indeed, take for granted.

HS: But do you folks have any kind of scheduled down time or times when people could expect network services to be down? Or do you really attempt to keep it up all the time?

MR: Well, we attempt to keep it up all the time, but there is a scheduled down time on Friday nights. This is largely reflective of students' schedules and people's personal schedules, when at least at this point, social life still tends to go on in non-Net-centric ways. And so we do actually have a scheduled PM time. Most of the services stay up during that time, but if we do need to schedule any kind of maintenance ,it's always scheduled into that particular window.

HS: Mark, one of the universities that an awful lot of people, when they talk about mobile computing, is Wake Forest, which I'm sure you're familiar with. And at Wake Forest, as I understand, all students, all incoming first year students are required to have a computer, and what they have is a particular kind of laptop that they get when they come in. Does Sonoma use that model, or are there other models that one could use? And could you talk about some of the different models and the pros and cons of how we get all students to have computers?

MR: Absolutely, and I think there are a number of dimensions here that are worth bringing out. In fact, the Sonoma model represents a very different polarity from the Wake Forest or Think Pad U model. This is not to say that one is necessarily better or worse. They are very different. Our objectives are very similar. David Brown at Wake Forest, who is behind their program, and I are very good friends, and we have, I think, very common goals in our respective programs. But we go about them in very different ways, at least in part because our institutions are very different. We are a public university. Wake Forest is a private institution. And both our financial and demographic characteristics are very different.

HS: Are you suggesting that private universities just have lots more money so the Sonoma model is less expensive?

MR: That's not necessarily the suggestion I was making, but that's certainly an attribute. But indeed there are parameters that one needs to take into account in public institutions that don't necessarily apply in private and vice versa. For example, our fees are set by legislative mandate, whereas Wake Forest could raise their fees.

HS: Which I believe they did, actually.

MR: . Which they did, and so for example, incorporating the cost of a program into the fee is something that Wake Forest can do but that many public universities could not. So--

HS: I just have a question about the fees really quickly here. So when Sonoma instituted this program, they did not raise their fees?

MR: Absolutely, because we in fact could not have done so if we'd wanted to. They are set by the legislature and they are standard across the system. That is true in many states, and even where they're not, the political challenges to raising fees in public institutions are usually a very great impediment. What we actually did, and the way I would characterize the two models, is a homogenous solution, a one-size-fits-all solution where the institution is very heavily engaged, where the financing runs through the institution in some way and where every student gets connected to the same machine, as opposed to the Sonoma model, which is a free market, student choice model where we set requirements and then it is up to the students to meet those requirements. There are a couple of other attributes that I think are worth noting. In one case, we're looking at fixed term with a refresh strategy, as Wake Forest has. Up to two years, they replace the machine. In our case, it's a varied lifetime and it's really dependent on what kind of machine the student bought in the first place. If they bought high up on the power curve, their machine will last longer and hence not need to be replaced as soon. If they bought at the bottom end of the power curve, they may need to replace more often. And that choice rests with the student. As long as they meet the minimum standards, they can choose to buy well beyond the minimum standards and thus get more life out of their machine.

JB: When you mentioned the standards, Mark, that's a question that certainly a lot of folks are thinking about. Just how do you, in this free market model and student choice model that you're describing that you use at Sonoma State, how do you communicate to the students what kind, what are those minimum requirements?

MR: Well, there are two ways of couching this. One is in terms of the hardware and configuration of the machine, and the other is in terms of capabilities. And we've taken the capabilities approach, although, of course, we need to then convert those capabilities into examples of hardware that meet them. And we've actually had the same baseline specifications since the very beginning of the program. I'll leave you to ponder what a specification might be that hasn't changed since 1995 and is still just as good as it was back then, and it's a one-liner. It says "Must be able to run the latest version of." It used to be Netscape. It's now Netscape or Internet Explorer. But nonetheless, the point is that what we said to our students is that we believe that the Web is very central to the way that we deliver things on the campus and hence, they must have the latest, greatest version of the Web browser. And so, given that they need to do that, in fact, the continuing increase in the hardware and software requirements for Web browsers has significantly changed what the actual hardware platform is from year to year. But the conceptual requirement remains the same.

JB: But that does--is likely to mean that students generally that came in four years ago probably have upgraded in some way, Mark. Is that fair to say?

MR: I think that is very fair to say. They are either upgrading their systems or trading their systems in for newer ones, but the point being that they opt when to do that, depending on what it is that they bought or what they already had, as in fact many cases we're hearing from our students, they already have machines that they'd like to bring.

HS: Now Mark, in the case of Wake Forest, Wake Forest actually buys the machine for the students?

MR: That's right.

HS: Right, the students don't pay a dime for the thing, at least in a way that the students recognize. But they're paying, it's buried in the fees essentially. What happens at Sonoma?

MR: The students buy the machines themselves.

HS: Are they subsidized at all?

MR: The are subsidized in various ways. In one instance, if a student is at the very low end of the income scale, is heavily maxed out on financial aid and has no other resources, we have a program under which we provide students free loaners. At the other extreme, the students who can go into the bookstore and charge it on their Mastercard or Visa, we do absolutely no subsidy for. In between, we have various programs that enable the students to bridge the gap between what they can afford and what they need through a work-for-a-computer program, through various kinds of favorable loan arrangements and so forth. And so what we try and do is to have a smorgasbord of options that match the students' various different purchasing abilities. Another model, by the way, is the one I just came back recently from Hong Kong, where Hong Kong University has implemented a variant of the Wake Forest model, where they have the one-size-fits-all Think Pad, but they've actually had the students buy it and have the university offer a subsidy towards the purchase. If the student gets through their four years of studies at the university, then they get to take the computer away with them and don't have to pay the university back. If they drop out--

HS: Yeah, they wind up taking away four-year-old computers that nobody would want anyway, but--Mark, you said the students have the ability to buy pretty much what they want, a free-market kind of thing. But are they buying from one manufacturer? Do you have a special deal with IBM or Dell or Compaq?

MR: We have a number of special deals, and that's another way we try and effect subsidies, is we effectively go out and we look for vendors that--the initial vendor with whom we had the deal for the first couple of years of the program was Apple. With the growing prominence of Win-tel machines, we now have deals with a number of other vendors. But the idea there is that those are deals that are negotiated based on the knowledge of the vendor that we have this requirement, and that they therefore come in as a partner in helping students meet the requirement.

HS: So you're still a mixed environment in terms of Apple and Win-tel?

MR: Absolutely.

HS: My understanding is that Wake Forest is not. Wake Forest is just Win-tel machines.

MR: Yes.

HS: Could you talk about how that works? I mean, it seems like it's simpler for the university to deal with all machines being the same. What's the downside or upside that you see of having this kind of mixed configuration?

MR: Intuitively, one would think that there was a huge downside in having a mix of different machines. In fact, the places where the standardization occurs is in things like office suites where we do have a standard that says you use Microsoft Office regardless of the platform. In the Web, in networking configurations and so forth. And these days, those things are sufficiently generic across platforms that we really have not run into a lot of support headaches around that.

HS: Do you have any sense of how many other institutions are out there that are doing this kind of thing, that are requiring people to have laptops?

MR: It's a large and growing number. The best thing I can say is dozens and dozens. It's probably under 100 still in the US, but it's certainly a significant number. It's growing rapidly. They range from, I think, at the high end of the largest institution that's committed to a wholesale program is the University of North Carolina and some other large institutions that are also committed. The University of Buffalo, SUNY and a couple of other relatively large institutions that are on board with these kinds of programs. University of Minnesota has made some very major commitments in that regard. At the other extreme are a lot of small, independent colleges that, for one reason or another, feel that this is something that they want to do and that the scale that they're working with is relatively manageable. Wake Forest runs an annual conference which I'll put in a plug for. It's happening, I believe, on the tenth of January this coming year, and it's a conference on ubiquitous computing access in which they invite all the people who are actually implementing such programs to get together and share ideas and so forth and there were a significant number at the last event, last January. And I understand that the attendance is doubling this year.

HS: You were there, at this last one?

MR: Yes, I was one of the (inaudible.)

HS: Do you have any sense that the one-size-fits-all model is more common or the free-market model is more common?

MR: At this point, I think it's fair to say the one-size-fits-all model is the more common one. It's got a great deal more marketing behind it, and I have to give a lot of credit to our friends at IBM for having made a real concerted effort to get people to understand what the characteristics of that model are and then to support it. And I think it's been a very effective marketing campaign on their part. In a sense, there's much less incentive for anybody to market the free-market model because there's no great money to be made for anybody, any one particular vendor or any one particular institution, in doing so. And so I think that in some ways, the one-size-fits-all model has been adopted more widely.

HS: Now, in all these cases, these universities are saying that we're going to do this for laptops. Nobody's doing this for desktops, at least the people you're talking about. There are actually some institutions that do either mixtures--in fact, Sonoma doesn't actually require--we strongly recommend, but we don't force students to get a laptop. We do have some students with desktops, although the proportion has been going down rapidly with the advent of ever more sophisticated laptops. And in fact, we're now starting to see the next step, which is that people are using a conventional laptop, a full size laptop as their non-portable, fixed machine, which sits at home or wherever and is only taken with them on rather lengthy ventures. And for their day-to-day carry-around, they're starting to carry around something smaller, a Libretto or a PDA or something like that.

HS: Yeah, actually, I will bet we start to see that with faculty and staff saying that a laptop is their machine, period, because I think they're getting so good. But when we look at people replacing desktops with laptops or just using laptops, it seems to me that we get a lot more than just a desktop machine that we can move around. Could you talk about some of the things that happen when lots of people have laptops, some of the advantages of universities having a big portion of their student body with laptops?

MR: Absolutely. I think first thing that you start seeing is that it is an empowering tool for students. It is very hard to be a passive user of a laptop. It forces the students to be active and, particularly as we look at what the students do with the machines, in every instance when I've talked to my colleagues and the surveys that we've done on our own institution and in general looking at what people use laptops for, they use them very largely as a collaborative and communications medium. And given that, the ability to collaborate around a dining table in the student cafeteria or send e-mail from anywhere or connect to the Web and share work with one another is something that really isn't enabled when you have a computer in a dorm room or even worse, as in our case. We are largely a non-residential institution, becoming more residential over time, by the way. But still, most of our students are off-campus. If they have a computer 30 miles away in their apartment, this really does absolutely nothing for them while they're on campus.

JB: When you describe that situation, Mark, about students, you know, with their laptops and getting around for meetings and working live while they're together like that, how do you, in fact, handle the--how do they access the network? And how have you solved the problem of the access to the network in that kind of a situation?

MR: We have some network access ports. From a technical perspective, what we've done is to create a campus-wide DHCP server that can dole out addresses so anybody coming in with a machine that has a standard DHCP configuration can simply plug it into a network jack and they get network dial tone, if you like, without having to configure anything. That's, by the way, a position we came to after a fair amount of experimentation with doing things in other ways. For quite a period of time, we had a registration scheme where any student that wanted to bring a laptop to campus had to first come in and register the hard address. And we then had a database of everybody's machines, and we did this out of a concern about security. There was a strong concern that, if anybody could just show up with a laptop, plug it in anywhere in the network, we would never be able to track the miscreants.

JB: Yeah.

MR: What it turned out was that this was actually a fairly significant amount of overhead, both in terms of running the program for us and for the students. And when the student would get a new machine, they'd have to come re-register it again. If an Ethernet card broke and they replaced it, then they'd have to re-register it. There were a lot of operational hurdles to making this smooth. In other words, there was a significant disincentive to plugging into the campus network by setting it up that way. We've now traded that for a little bit more work. If we have an abusive incident of some kind, there are ways of tracking the abuser because typically people don't just come in, bring in a machine, use it once and then go away and never bring it back. And so there certainly are ways using sniffers and so forth to actually get to which particular hard address is associated with a particular event. But it just takes a little bit longer. It does put more work into the tracking of security incidents, but at the same time, it's made the system much friendlier and much easier for both us and the students to use.

JB: In other words, you found that you could really decrease the amount of support that was necessary to support this universal access, then, by not requiring that hard registration of the computers?

MR: Right, and we've increased our workload somewhat in tracking security incidents, but it's been a good trade-off.

JB: We do have a question that came in from Mike Abney that links--that kind of addresses some of the support questions, and he's asking what type of ratio of notebooks to helpdesk support staff would you recommend is needed to adequately support the laptops? That probably raises lots of questions, doesn't it?

MR: Absolutely does. Really, very much depends on the model that you are using. If you have a heavily institutionally committed model, where the institution is directly dealing with maintenance of machines, where it's dealing with provision of dial-up services and so forth, potentially you have a huge support burden. If you rely on a degree of student independence on this--in our case, for example, we've decided not support, with some honorable exceptions, but in general, we do not provide Internet service access for dial-up modems. We do have a pool of modems for faculty and staff which we maintain, sort of the traditional university model, but other than that, we typically tell our students, "Go out and get an account from a commercial Internet service provider." This works really well in any part of the country where you have good competition. In our case, last I looked, we had 32 Internet service providers with local numbers in our area. Obviously, if you live in the remotest part of some remote state, this doesn't work as well and you may have to create that particular service. One of the interesting ways in which this ties back, though, to the question of ratios of helpdesk staff to machines is that when we did an analysis of what our helpdesk was doing, in the early days when we did provide dial-up services, that 80% of all the calls that we were managing had to do in some way with problems with dial-up. So we could cut down literally by to one-fifth of the service support that we--

HS: Eighty percent of your helpdesk disappeared?

MR: . No, we just started--for the first time, when we got rid of it, we started having an effective helpdesk. It was way understaffed as long as we were handling dial-up.

HS: Mark, so you're saying that if staff members and faculty members who want to do computing from home, which is--a lot of staff members and faculty members at least that I'm familiar with do that--you're saying that they pay their own ISP fees?

MR: No, in our case, what we've done is to say if students want to do it, they pay their own ISP fees. If faculty and staff want to do it, we have a residual pool. What we basically did was to take our existing modem pool and to carve it up. We gave a piece of it to faculty and staff, and that continues to be free. And then we gave a piece of it to the neediest students, and again, it's a needs-based thing. Financial Aid office basically authorizes us to give out a free ISP account to any student that can prove appropriate financial hardship, and so again, the idea is rely on the students to pay for it but provide a safety net. Very similar approach to what we did with the computers themselves. What happens with the faculty in our case, for example, about three quarters of our faculty do not live within local dial range of the campus. So what we say is it's free for anyone, but if you live 30 miles away, it'' a long distance call. It's probably a lot cheaper to sign up with an ISP that has a local number. And so that's a way of constraining somewhat the number of people who actually use the campus dial-up facility.

HS: But these folks coming in from ISP's, can they get in and get access to your Novell network, if you have such a thing? Or can they run IMAP mail servers from there? I mean, can they really get all the services they need or you just really give them kind of Web access?

MR: Well, no. We give them all the services that they need, and in fact, that's taken a fair amount of network engineering. I'm thinking back to the topic of today's talk, which is really networking for this kind of access. One of the things that we had to do was to create proxy services for a lot of services to which we subscribe, which are validated through host IP validation. In other words, if we subscribe to, for example, a commercial database provider within the library and the provider says this can only be used by machines originating in the Sonoma.edu domain, what we have to do is to place a proxy server in front of that service which then validates, authenticates the students. Every student receives a Kerberos ID, and so basically, when they come in, they can get in either by just being physically on the campus and being host IP validated, or they can get validated through the proxy server and at that point, then, the proxy server is the entity that's actually talking to the database provider or whatever. In answer to your question about Novell, network, no, it doesn't. But we long ago decided that we were going to be a single protocol network, and to the degree that there is Novell or Apple Talk left on the campus, it's local to specific environments, and we don't propagate it at a distance.

HS: Mark, one of the areas that people think about, and I think increasingly think about when they think about mobile computing today, is wireless mobile computing. Are you doing any of that at Sonoma, and can you talk a little bit about if you are, how's that working? And if you're not, why not?

MR: We've looked at it. We're not at this point doing any production wireless networking, but in fact, somewhat driven by the fact that I was going to need to do this presentation, I decided to do a statewide survey in California a little while ago to find out what our colleagues at the other Cal State campuses and the community colleges and elsewhere in education in California were doing. And what I learned was that, in fact, very few of them are doing anything very substantial with wireless LANs. First of all, when you look at wireless networking, there are two kinds of technologies. There are point-to-point wireless technologies, which are pretty mature at this point. They can actually do pretty high bandwidth links. Ten megabits is quite regular, and there are actually even DS 3 speed links that you can buy. They're essentially ways of bridging buildings across roads and so forth. A number of places around the state are using those. They're particularly useful in the city settings where you want to bypass the need to dig through streets and so forth. The technologies that I think you're asking me about are wireless LAN technologies, and in looking at them, there are a couple of instances in California of two or three different technologies. And I just want to run through those briefly to give you an idea of what's out there. The high end is something, this company called Radio LAN. They have a ten megabit system. It's an unlicensed five gigaHertz system. Each of the base units covers about 120-foot radius in a semi-open space. They're really thinking of this as a technology for cubicles and so forth in office spaces. So about 10,000 square feet per repeater. Another technology comes, something called BreezeNet. This is a shared three-megabit technology, and the senders--we actually have a couple of campuses, Hayward and San Marcos, that are experimenting with that on a limited scale right now. The three-megabit shared between anything up to eight users on any given repeater. The technology there, it's about 1,500 bucks for a base unit and $500 for a PC card. There is a Metricom option. This his been around for quite a while. It's low bandwidth. I'm sort of going in decreasing bandwidth order. That's a 38.4 shared bandwidth. But the nice thing about it is that Metricom puts in the infrastructure, and they typically extend it not just over a campus, but also out into surrounding communities, up on the lampposts and so forth. And there, the costs are basically just to the student. There are no institutional support costs, but students pay either $20 a month and buy a modem or $30 a month and get a modem included. Metricom has been selling that in a number of institutions through their bookstores and so forth. So it gives you an idea of sort of the range of options that are out there. Now, you asked how well these things work. Key issues, really, for us are that right now on the wired side, we're seeing people starting to move from ten megabits to 100 megabits and thinking that ten megabits isn't quite fast enough. And so bandwidth is an ongoing problem with these things, particularly when you're in an environment where you have shared bandwidth, as in the case of the--

HS: But you see, I mean, are there technical limitations or is this just a matter of give it a little time and they'll get there?

MR: Well, I mean, one of the sort of the absolute technical limitations is that you can constrain a lot of bandwidth inside a fiber optic or inside a copper wire. When it's out there in the air, there's a shared bandwidth, and so there's obviously less bandwidth in airborne media.

HS: Sure, less bandwidth in a chunk of air than if you ran a thousand fibers through (inaudible.)

MR: Exactly. So that's a physical limitation. There are some power issues. One of the issues, obviously, is if you put a transmitter into a laptop, and battery lifetime in laptops is always a problem anyway, if you add a fairly powerful transmitter as these things typically need to have, you significantly reduce the battery lifetime from what is in many cases already a rather short battery lifetime. And the sorts of issues that we've come up against and the thing that actually stopped us from doing any work on the campus with it is that there are real problems if you look at the typical campus setting, where a lot of us have concrete buildings with steel reinforcement and so forth. What we've typically found is that there are lots of dead spots. It's very hard to get really good coverage in a typical institutional setting.

HS: Actually, you have both problems. You have a problem of you want to put in something hard-wired, you've got to chip through all that concrete and steel, making that difficult. And you're saying that the very stuff that would make you want to put wireless in makes it difficult to transfer.

MR: Exactly so. At the same time, what happens typically--and all of us who've used cell phones, I think, will empathize with this--even when you actually think you've got it covered, strange artifacts arise and you have a degree of flakiness. I mean, there is no way I would have given this presentation on a cell phone because I wouldn't have been able to trust it. And this comes back to your very first-ever question, which is--

HS: Just stay out of tunnels, Mark, and it will be okay.

MR: . Well, even then, don't count on it. I think anybody who uses a cell phone will know that you cannot reliably count on it. If you want to have an enterprise-grade network, which we need because we're relying on it, wireless is at this point still a long way from being enterprise grade. It is a useful technology. It is a useful direction to go in, but the surprise to me in doing this survey was how little adoption of wireless I found. I mean, I could not find a single institution in California that had made a strategic, institution-wide commitment to wireless networking.

HS: It will be interesting to see if that's still true a year from now. I mean, it will be interesting to look at the acceleration of this technology, whether it's flat or it really takes off.

MR: Right.

HS: You think it's going to take off?

MR: My sense is that it will continue to grow in a relatively controlled way. I don't see it replacing wired technologies, particularly with the increasing bandwidth that we're seeing. When you look at something like Internet2 and what it's doing for national bandwidth, it's very hard to see that people would want to be so constrained on the end point connection.

JB: We did have a question coming in from Bill Klinko at Cornell, asking about the advantages and disadvantages of wireless, and I think you've addressed some of those questions, Mark. But then Bill also was asking about the costs involved in those wireless ones, so if you could address that. And also whether you've got any experience in terms of how this extends to palmtop technology.

MR: . Well, that actually raises another wireless technology I didn't mention, which is short infrared, which at this point everything is equipped with. And I've had some interesting observations. These are anecdotal, but I've seen this replicated in enough settings that I think it's worth sharing. In the work that I do, I find myself quite often sitting down with vendors and doing various kinds of focus groups and other sessions on requirements for new hardware devices and so forth. And all of the vendors have convinced themselves that they have to put infrared capability on everything. And then when they, in any focus group that I've ever been at, ask the question, "How many of you are actually using it right now?" no hands go up. It's a very pervasive technology which is out there in many, many settings, and very few of us use it. The one honorable exception to that is in palmtop links. It does appear that people who have palmtops are starting to use infrared to connect back to their home machine, whatever it is, the laptop or the desktop.

JB: So they're not communicating with somebody else. They're communicating with themselves, from themselves?

MR: Yes. I mean, we would love to see--the concept is wonderful, that you could walk around with this thing and just sort of walk into a room and point your machine at somebody else and communicate with them. I have yet to meet anyone who actually does that on any kind of large scale or any institution that does it on a large scale.

HS: Obviously, infrared is much worse than RF in that it doesn't go through walls at all or doors or anything, even the palm of your hand. I mean, anything will stop it.

MR: Right. It's also potentially much better in that regard because you now no longer have the problem of limited bandwidth.

HS: That's true.

MR: It's an encapsulated--

HS: Yeah. When I opened, I said something that I knew you disagreed with, but I said it on purpose. And namely, I talked about wiring every classroom and every seat, which is something I hear all the time. When we talked to you on Tuesday, you said that might not be the best use of technology. Could you talk about that? Because I think a lot of people believe that if they're going to do mobile computing, that's the direction they're going to go it.

MR: Yes. The pretty picture is one in which you have every single classroom wired and lots of students sitting, with rows of computers, facing the front of the room. And when I see that, I'm reminded of what people did when the movie camera was first invented. They thought a movie was a stage play, and so they set up a camera in front of a stage and then they had a play, and they filmed it and they said, "This is a movie." And not until we started changing the very paradigms for what's represented did movies take off, because in fact, a real stage play is a lot better than a movie of a stage play. And so at least in part, some of the reactions, I think--negative reactions people have to technology in the classroom is that they're thinking of technology in the classroom!

JB: Mark, does that mean that, you know, if I'm doing strategic planning for a campus, that the goal--we don't necessarily have to be thinking about a goal of having a network connection available in the classroom, every classroom for every student?

MR: I think absolutely not. I think you need to have some classrooms with some network connections available.

HS: What about a network connection for every faculty member in every classroom? I mean, we would probably agree--would we?--that every professor walking into every classroom should be able have access to the network.

MR: Yes, and I'd take that one step further and suggest that there should be access in every classroom for the network because it may be not that the professor wants to use it, but that the student wants to get up and show something and use it in that context.

HS: Okay, but that also implies that not only--I mean, this takes this a step away from network computing, but if the professor's going to come in and show it, it says we really need not only network connections, but some projection facility in every classroom.

MR: Or at least in a significant number of classrooms. Yes, that's been our experience. That's one of the most heavily desired resources. It's by the way used at least as much by students as it is by professors.

HS: And your idea is not that there be a computer anchored into this classroom, but that the professor would bring his or her own computer in, plug it in and project it up on the wall. Or a student would bring their own computer in and project it up on the wall or a screen, whatever.

MR: What happens today in "the real world," people walk in with laptops and give presentations. If you go to any conference today, very few people walk into the room expecting to see a computer wired up to a projector. They'll walk in with their laptop, plug it in and present.

HS: Okay, if we're not going to have these connections, these remote connections at every seat in every classroom, could you talk about some of the places we ought to have them, besides classrooms, we ought to have these connections?

MR: We absolutely need to provide work spaces where students can get together and either independently or collaboratively work on projects and have access to the network. And the obvious place is--and many places have taken this direction--are in residence hall rooms. So the one port per pillow idea is, in fact, a very popular one these days, and that's somewhat self-evident. Other places that are fairly self-evident are places like library carrels and study rooms and study spaces. Perhaps less self-evident but still very important are social spaces of various kinds. I mentioned earlier on students sitting around a laptop over lunch, working on something, so certainly having access, for example, in a student cafeteria is something to think about. And I don't know that I want to go on the record--I guess this is going to be recording, so I'm embarrassed to say this, but the ultimate fanciful version of this is ten base tree. You plug into the bottom of a tree trunk outside, and assuming that you can see the screen--which is still a problem--then you have the ability to compute wherever and whenever you want.

HS: But you don't need to plug into the tree if wireless would work, right?

MR: That's right, and it may very well be that there are settings where wireless is adequate.

HS: I can imagine telling our folks here that we have to wire every tree on campus.

JB: Maybe we have the towers are attached to the trees for the wireless transmitters?

MR: You have to forgive me for being a little fanciful here.

HS: But we like the idea.

MR: But the point is that technology does not just live in classroom. In fact, the hopes--the key answer to Judith's question here is that rather than trying to increase the use of technology in the classroom, let's decrease the use of the classroom. The classroom is a great tool for traditional instruction. It is not a great tool for technology instruction.

JB: I think that's probably a very interesting point to have you suggesting, then, that we really take a look at our campus, Mark, and perhaps really think about that new paradigm and where the networking is going to be used, then, is what you're suggesting.

MR: And I think there is possibly a trade-off between physical plant and computer networks to the degree that you can get learning to happen in non-classroom settings. And if you have increasing enrollment, as many of us do, you may not have to build as many new classrooms if you can have more of the learning take place outside the classroom.

JB: Well, that's an interesting thought. You know, we're running close to the end of our session here and we've got a number of questions that we haven't gotten to. But Mark or Howard, is there anything else, a final comment or question that you'd like to--

HS: Actually, I'd like to get one last question in here. I fear that I'm going to ask this and Mark's going to talk for an hour on it.

JB: We'll stop him.

HS: But I think that a point we should address before closing is that there's lots of universities and colleges out there who really haven't gotten started on this road to mobile network computing. Could you talk briefly about how they might get going on this kind of thing? What's it going to take?

MR: I'm very glad you asked because that's the last issue I was going to raise. If you look at these kinds of initiatives, I think that what you need to recognize is that they are institution-wide initiatives. They are not IT initiatives. They don't just affect the IT organization. They require a great deal of pre-planning, buy-in and considered action by a range of entities around a campus. The bookstore, the financial aid office, the admissions office, the library, the faculty, the senior administration have to recognize that they're transforming their institution, not just bringing in computers. And all of these things need to happen in parallel with other kinds of initiatives. You can't just put computers in the hands of students without creating the network infrastructure to support them, without creating the support services and the helpdesk, without creating the online information resources within the library, and without creating the faculty development and other kinds of programmatic activities that, in fact, are both dependent upon the laptop initiative but also are ultimately the reason for doing it.

HS: Okay, I think we're just about out of time now, Judith, are we?

JB: I think we probably are, and I think we'll just tell our many viewers and listeners to watch the Website for follow-up on some of the other questions that we didn't get to. We had a number of questions that weren't directly on the mobile networking, so watch for those at the Website. With that, I'd like to just say thank you to everyone for participating with us today on this session with Mark, and to go ahead and also send--follow-up questions can be sent for the next couple of days yet to expert@cren.net. Also, be sure and mark your calendars for the new spring series launching on January 14, 1999. The January 14 Tech Talk will feature two experts from Cornell University, Steve Verona, technologist, and Margie Hodges, a legal expert. These two experts will discuss issues on intellectual property rights in the Web environment. So please join us for that. Also, do check the Website for a full schedule of what we have set up for the spring. And the full schedule will be posted as it gets firmed up. We also welcome suggestions and feedback on what you'd like to see and hear here at Tech Talk. Thanks to everyone who helped make this possible today: the board of CREN; our guest expert, Mark Resmer; technology anchor Howard Strauss; and to Paul Bennett at UM Web Services for the encoding. Thanks, too, to all of you for being here. You were here because it's time. 'Bye, Mark.

MR: 'Bye-bye. Thank you very much.

JB: Okay. 'Bye, Howard.

HS: 'Bye, Judith. 'Bye, Mark.

JB: 'Bye, everyone.

[Top of Page]