The magazine of the Melbourne PC User Group A Brave New Digital World Mike Chambers

One of the great dangers of prognostication is the trap of linear thinking. We can look at today's technology and products plans and extrapolate a future device that is commensurately faster and cheaper. However, this linear view of the future is often upset by disruptive events and technologies, driven by the changing ways in which technology is used. Disruptive technologies have given us jet airplanes instead of cross-country passenger trains, broadband Networks in place of marginally faster analogue communications.

The origin of disruptive technology is interesting as well. Historians ask, Does the man make history or does history make the man? As applied here, Is our technological future shaped by our technological innovations or by a vacuum of needs that breakthrough technology must fill? I think the answer is "Yes" to both.

The notion of utility computing is not a new one. Technologists and marketers have been spouting about it for decades. What is new, I think, is the understanding that utility computing will reshape not only the way in which computing is bought and sold, but also the very devices that we personally compute with. Let's look at ten trends that I believe are shaping the future devices we will call personal computers.

Trend #1: The Changing Nature of the 21st Century Workplace

Increasingly, corporations around the world are not only embracing telecommuting - they are now insisting on it. Larger numbers of employees are being asked to share space while in the office in guest or traveller cubicles that feature a phone and a LAN connection. Offices are being streamlined and downsized. Customer-facing employees are expected to spend the bulk of their time facing customers, not sitting at a desk. And even those employees chained to a desk (like call centre and support desk workers) are increasingly sharing desk space by shift, as more corporations seek to provide 24-hour service to a global marketplace. In many cities, the shift has another motive: overcrowding and pollution that accompany massive commuting patterns are being addressed by mandated reductions in the number of vehicles coming into the office car park.

For these and other reasons, the notion that one has a personal space at the office is rapidly disappearing. Instead, the mobile workforce is staking out a cyber home on the global information superhighway. The implication is that my computing and communications environment becomes my virtual personal space. The electronic desktop metaphor becomes an increasingly powerful one.

My windows desktop bitmap showcases the family photos that once sat in a frame on my credenza. That family today may be minutes away and tomorrow may be halfway across the country or across the world, as jobs become more global and more mobile. Web portal home pages give me a virtual window into my cyber world, tailored to my interests and tastes. They watch my weather, finances, travel plans, impending wars (military and cybernetic), and the latest news. It's my morning paper, minus the paper, updated 24-hours a day.

Trend #2: The Nature of Work is changing as well

Gone are the days of salesmen courting clients with two hour, three martini lunches. Business today is moving at Internet speed. Technology is both driving this change AND is driven by it. Digital interactions (like Webinars and NetMeetings) are growing as a means of rapidly communicating complex ideas to clients and coworkers effectively and efficiently. These tools are allowing global teaming and collaboration to happen on a much broader scale than ever before.

Increasingly, the nature of work requires integrated voice, data, and visual communications that can be used at a moment's notice.

In addition, larger and larger proportions of the work force are becoming information workers - people for whom voice and/or data Networks are an indispensable tool.

Trend #3: The convergence of mobile technologies

The past year has already seen the first phase of convergence, as PDAs and mobile phones took their first awkward steps toward unity. Don't let the imperfection of these first steps mislead you: these devices must be united. Too much of what PDAs store are the people and companies we must stay in touch with. Mobile phones give us the medium for making that contact. Technological nature abhors the vacuum epitomized by looking a number up on one device and dialling it on another. These devices will converge and the world will be a better place for it.

With disk sizes exceeding 20 GB on laptops and 200 GB on desktops, this seems implausible. However the OS, applications, and temporary storage consume the vast majority of storage used on typical corporate PCs.

Microsoft's Pocket PC 2002 demonstrates beautifully that bloatware can go on a diet by fitting the majority of the Microsoft Office Suite functionality and an operating system into a 64 MB Flash ROM.

Perhaps sad, but very true: 99.9% of what I do on a PC or Laptop everyday, I could do on my Pocket PC with the right ergonomic interface.

Today, the limiting factor on these devices is an artificial one: memory size. Today's Pocket PCs sport 32 - 64 MB of user memory, which is split between program execution space and file storage space. While this is a lot of space, it's not enough to keep up with corporate e-mail and the 2 MB attachments I see all too frequently.

Flash memory densities continue to double about every 14 months. As consumers and manufacturers have slowly come to grips with non-volatile storage, devices of enormous importance have emerged. The 32 and 64 MB USB flash disks were interesting, cute and occasionally handy to move a large file or two without a Network. But as capacities have grown to 128, 256, 512 MB and 1 GB, the very real possibility exists to keep ALL of one's data on a non-volatile flash disk the size of key chain ornament. Lexar has recently announced a 4 GB compact flash card, pointing way toward very high density, very high capacity NVRAM devices changing the face of personal storage.

As NVRAM capacities grow and prices plummet, system makers (of everything from PDAs to laptops to desktops) will be forced to wrestle with how these device much change to embrace and fully exploit the storage medium.

Look for next generation palm devices to incorporate over a gigabyte of NVRAM in the not too distant future, while storage card sizes continue to double. These devices already hold great promise as the personal consolidation platform for the electronic world: PDA, Cell phone, MP3 Player, and even video playback are possible today. Tomorrow's technology will scale these devices to challenge the need for a laptop.

Trend #4: Net Apps change everything else

Of course, not everything happens inside Microsoft Office. Corporate applications like SAP, PeopleSoft, Siebel and others will continue to grow, and mobile professionals will still have to use them. Increasingly though, the user interface to these applications will be browser based, reducing the need for high power workstations to handle client-side enterprise applications.

Along with this trend, I think we'll see another disruptive change in the natural order of application software: real-time delivery of application functionality. The fact of the matter is that I need less than 10% of my office-suite 99% of the time. Much of the rest, I'll never (or very rarely) use. We've become fairly accustomed to installing on-demand functions within our browsers, and I think we'll see much more of that in our future applications.

This promises to change not only the way applications are written and installed, but potentially the way that they are sold. In essence, applications can become like long distance telephone services. You don't pay to have a continuous telephone connection to every spot in the world that you might someday call. You pay for the bandwidth when and where you need it. Pervasive Networking (wired and wireless) will allow pay-per use application functionality. Need to create a PDF file? A 25 cent charge shows up on your account. This is one of those bizarre business evolutions that is as good for the consumer as it is for the supplier. A lot more PDF files will get created when it doesn't require buying and installing a massive package. The consumer gets functionality when and where he needs it. More consumers will then get access to the functionality they need.
The application vendor gets broader use of his functionality by a broader user base and thus larger revenues.

Perhaps you think this pie in the sky dream is way out there. Perhaps you are right - but perhaps not.
Microsoft's recently announced Dynamic Systems Initiative promises to rethink the way that systems and application are deployed, going all the way back to the development architecture. Much of the application execution model we have today is legacy inheritance from 40-year-old batch mainframe architectures. DSI holds out the hope of rethinking how systems and applications are developed and delivered to end-users. Portions of the DSI development model will be included in the .NET Developer's Kit. The System Definition Model (SDM) and Automated Deployment System (ADS) offer developers a framework for flexible runtime functionality deployment that delivers the functionality the user wants when he needs it and on the platform he needs it on.

Trend #5: PC² Devices drive Converged Network Terminals.

The constituent words here have connotations that may cloud the meaning, so let's get very specific. A Network terminal is an interface where a Network device may be connected to the Network. It is converged in the sense that we no longer will see a CRT and a phone (a data terminal and voice terminal respectively). Instead these devices will converge into a single Network terminal.

By analogy, the power receptacle your computer is plugged into is a Network terminal as well. It's the point at which various electrical devices can connect to the electrical Network that powers them.

Terminal also denotes that these are not especially sophisticated devices. Their job is to connect the user to the Network: user interface. The intelligence remains in the PC² devices that connect to the terminal. This separation of functions allows the common, heavy components of an ergonomic interface to become stationary - like the desk phone in guest cubicle - while the personalized, scalable and more upgradeable functions (processor, storage) go with the individual.

If you think about it, there are some compelling reasons that corporations would like to see these devices. Chances are you've heard a coworker pleading the case for a better PC. But have you ever heard a co-worker whining to the boss, I just can't take this telephone anymore - I've just got to have a new one. A phone's a phone. It's a voice terminal. It provides the human user interface to the voice Network. Likewise when mainframes were the rage, there wasn't that much yammering about getting the latest and greatest data terminal. The computers behind the terminal could come and go, upgrade, get distributed, and host new applications. But the terminal did its job year in and year out, providing the human user interface to the data Network.

Trend #7: CNTs & PC² devices in the schools

In addition, expect to see PC² devices connected to CNTs emerging in the classroom, where they will offer new dimensions to the learning environment and the way in which teachers, students, and parents interact and collaborate in the education process. Integration will mean that assignments appear as tasks on a student's PIM calendar, work is electronically turned in, sometimes graded automatically, and results communicated digitally.

Of all the trends I see, this is perhaps the most encouraging. We have the opportunity to educate our children in ways which they enjoy and use technology to provide a level of one-on-one interaction in the learning process that is simply impossible today. This trend will be a blessing to parents of learning-challenged, who will be able to easily track daily the progress and obstacles in their children's education. Additionally, it will be a godsend to the parents of gifted children, who can be tackling problems at their own level. I believe that educational software itself will become more heuristic and tailored to the natural human learning processes.

Far-fetched, you say? Think again. Japan's Broadband School Initiative, supported by Microsoft and 18 international technology firms, is deploying wireless broadband connectivity to every school in Japan, starting today with 13 Tokyo elementary and middle schools. Notebook computers wireless connected to the school's broadband Network will provide educational content to students and better interaction for teachers and parents.

Trend #8: Pervasive Wireless Broadband.

It won't be easy getting there, but this will happen. As I originally wrote this piece I said, By the end of this year, we will see the kinks and incompatibilities worked out, and 802.11g wireless access points will own the airwaves for a time. Intel's Centrino announcement is giving me pause at this point. I believe the kinks will get worked out of 802.11g, and it should be the wireless platform standard. However, Intel's decision to standardise Centrino on 802.11b and the massive launch of the platform by major computing vendors blows a little smoke over the crystal ball. Nevertheless, whether by "b" or "g" we are now entering the age of the wireless world.

Frankly, I think statesmen should wake up to the notion that this is truly the information superhighway. It is a piece of national infrastructure that needs rapid development and deployment. It is every bit as important as national transit. Though I am normally wary of asking government to do anything that private industry may, in this case I think government construction of the wireless superhighway makes every bit as much sense as government construction of our roads and bridges.

I am more concerned about the fragmented, exploitative commercialisation of wireless access emerging now. In the U.S. for example, McDonald's is adding wireless access to 1,100 of its stores, where you may pay by the day or the month for Network usage. (In one stop you can get both your bits and your bites.) So what does that ultimately look like? A subscription to McDonald's wireless for lunch browsing, Starbuck's wireless for the morning news, and so on?

Ultimately, wireless broadband needs to be seen as an essential element of a nation's infrastructure. Without doubt, there are both conceptual and technological hurdles to meet and competing thoughts about how to meet them. What we cannot afford is to smother the emerging capability with hordes of opportunists creating a quagmire of overlapping, incompatible, and costly services.

Really the question becomes How should a nation pay for wireless infrastructure and how much profit should be made from the undertaking? Is it sensible for a handful of corporations to own the national highway system? If not, how is it sensible that they should own the wireless information highway?

These socio-political questions notwithstanding, ubiquitous wireless broadband is coming soon.

Trend #9: Broadband evolution

If wireless broadband is to become pervasive, current broadband offerings must change or die. Frankly, I'm tempted to say that either suits me fine.

If nationwide wireless access becomes as easy as mobile phone coverage, current broadband offerings must find a way to differentiate themselves. This could happen in a number of ways.

First (and perhaps most desirable) would be 100 Mbit & Gigabit connections. DSL will be architecturally challenged to do this, as the link speed limit is inversely proportional to the distance from the central office. Most of the population does not live close enough even to get today's maximum throughput of 1500 Kbits/s. Meanwhile cable modem subscribers routinely get 2200 Kbits/s, with some achieving much higher 7,000 Kbit/s -close to the theoretical maximum for a 10 Mbit.

Perhaps more likely will be a long needed improvement to basic service. Broadband ISPs typically offer 1 to 5 e-mail addresses and 5 or 10 MB of Web storage. Meanwhile hosting services offer hundreds of megabytes of storage and unlimited e-mail accounts and advance features for a few dollars a month. Broadband ISPs can and should provide personal hosting accounts that eliminate the need for the Web hosting services. And while they are working this, performance tuning on the hosting side is desperately needed.

Trend #10: Device Driven Security, pervasive VPNs

Again, the foreshadowing of this trend is here today. Many corporations now use VPN technology to allow mobile employees a secure tunnel into the corporate Network while using the public Network.

The inevitable growth in wireless intranet access points creates an enormous security "oops". Network designers have rather depended on physical Network access to determine whether you were an internal, intranet users or an external, Internet user. Wireless access points explode that assumption. IT managers are already battling rogue access points hidden on their Network. These devices are a potential security nightmare because they allow a route to bypass the firewall without physical access to the Network premises. Given the widespread availability of wireless access point technology and its ease of deployment, managers of large scale Networks should assume as a matter of probability that they have rogue access points today and from this point forward.

Solving this problem is much harder than pointing it out, and I don't suggest that I have the answer to this challenge. However, I suspect that IT managers may be forced to a Network architecture which assumes that all user logons are external and thus require all logins to use VPN tunneling or some other means of encryption and validation.

Clearly though, future computing devices will contain much larger amounts of information about the people using them.

Bonus Trend: Divergence at Home and Office

With all the emphasis here on wireless and mobility, is there a future in stay-put computing? I think so. However, I think that we may see a trend away from PC based gaming as dedicated function gaming systems become more capable, widespread and intelligent. Look also for Network attached storage to grow in popularity in the consumer market, as it makes increasing sense to have a central home storage repository for large scale data as more and more personal devices go solid state.

Conclusion

While component manufacturers are making nearly unbelievable strides in the delivery of high performance, low cost parts for personal computers, it's less clear that these advances will be what sells the next generation PC. Where more power has been the mantra of the past, low power may be the calling card of the future (as we've seen just recently with Intel's announcement of the "Pentium M" processor and the Centrino mobile architecture).

Will this possible future of personal computing really come to pass? I don't believe that technological possibility will hold it back. The largest obstacles are the technology providers themselves. Historically, technology providers have created and provided more capacity and power than the consumer needs, and they might choose to continue to do so. They must come to understand that leading edge and state of the art may no longer be measured in gigahertz, megabytes, or terabytes. Rather, true industry leadership may need to be measured in how well emerging technologies fit and enable the trends that shape the way the 21st century world computes.