Most Helpful Customer Reviews
|
|
29 of 31 people found the following review helpful:
5.0 out of 5 stars
A Vision for Designing More Useful Information Technology, January 19, 2001
Although this book was written for both people who use computers and for the technologists who use them, the latter are the primary audience. General computer users will find their normal complaints about bulky, balky technology recognized here, but will get little but emotional support for near-term improvements. The primary benefit of the book comes in the many scenarios of interactions with information technology to simplify, speed, ease, and improve the processing to better serve the user's needs.Dr. Dertouzos is always on the cutting edge of the information revolution in his role as the head of MIT's Computer Laboratory. The core of this book is captured in chapter 8, where MIT's new Oxygen project is described. This is a prototype of "human-centered" information technology. The system combines a portable device for wireless communication, a stationary system built into a room (with transportable software from the portable device to the stationary system), and a network to support the interactions of users to the technology in new ways. The strongest part of the book is in complaints about the limitations of current information devices and networks. These will be familiar to any computer user, but it is refreshing to hear them from someone involved in drawing the outlines of the future. These include bulky software that does too much (like the word processing program most of us use that keeps automatically reformating what you have typed into something you don't want), weak interfaces between multiple programs and products so they crash when combined, the need to type so much information in, lousy search engines that waste your time, horrible telephone robots for getting to the right number, difficulties in sharing information, and the burdens of unwanted and unneeded e-mail. His solutions focus on five areas: Letting people converse with information devices in ways similar to how you would speak with a service person in a business; using e-forms to capture your information once and to then automate the sharing of that information with organizations who legitimately need it; finding answers by building on information that others have learned whom you trust; changing the method of distance working and learning so that the interactions are made more realistic and better summarized; and allowing you to tap into personalized, custom software preferences wherever you are and with whatever device you are using. Each area contains several examples of how these changes might work, many drawn from actual Oxygen applications that are now operating. So you should think of this book as focusing on what will be technically feasible in the next five years or so. I hope that Dr. Dertouzos will write a sequel to this book that looks further ahead than that in order to begin to spell out an even more improved version of information processing. As much as I was attracted to his vision here, I found that it mainly focused on enhancing the ways that I do things now. I thought that more could be done to help individuals operate in new ways that would vastly enhance human progress. Problem-solving software designed to help structure issues, gather information, analyze it, get feedback from others on the process, and compare to the potential for perfection could be one such example. Seeing this book also made me realize that much more work of this sort is needed. Without detailed scenarios of how to create solutions that people really want, technologists will continue to provide user unfriendly technology. I suspect that we need a vast experimental activity where people attempt to find new ways to get benefits from technology while removing its hindrances. Those who read about "human-centered" technology will, of course, want to know what the catch is. You will find towards the end of the book that Dr. Dertouzos points out that making the humans a little more standard in their interactions would allow the information technology to work better. So the vision is still a little along the lines of making each of us fit into the round hole in the technology board. With more technology advances, I hope that aspect will quickly disappear. It certainly should be a primary objective. After you finish reading this book, I suggest that you create your own scenario for a better way to get a task done with information technology. Then send it along to Dr. Dertouzos, so he can share it with others. In that way, you can help speed the unfinished revolution talked about in this book. Let's focus on making vast improvements in human benefits, net of human frustration and stress, in all of our technologies rather than focusing on selling products to other technologists! That's the real mindframe shift that is needed!
|
|
|
20 of 24 people found the following review helpful:
4.0 out of 5 stars
Unfinishable revolution?, April 25, 2001
Few people have more credentials to speak about progress and challenges in human-computer interaction than M. Dertouzos - the head of the Laboratory of Computer Sciences at MIT, which has a distinguished record of cutting-edge research in this and other fields. And yet many arguments and predictions in the book remain somewhat unconvincing."Why computers aren't as easy to use as cars?" - asks the author, like many other people before him, frustrated by their perpetual complexity and cumbersomeness. But comparison with cars is misleading. Cars are not designed to allow motorists to put under the hood any additional gadgets they fancy, or to perform arbitrary maneuvers, pushing every button and handle simultaneously. Yet the development of PC industry was based on accommodating ever more and newer gadgets under its cover, and on allowing almost any user's action, short of whacking a motherboard with a sledgehammer. Of course, many flaws of computer systems are due to the industry's geeky origins and traditions, or specific biases of programmers and early users. But the roadmap described by the author is not the first serious attempt at radical improvement, and the goal it is hardly closer today than a few years ago. Why? This probably has a lot to do with the economics of the computer industry, rather than other, more subjective, factors. As much as both hardware and software companies try to convince us how hard are they working to improve usability of their products, to eliminate bugs and crashes, the dirty secret of the industry is that it is not a top priority. Quality simply does not pay. In the "physical" world we often buy new things just to replace broken, or worn-out ones, not necessarily because the older items are hopelessly obsolete. Manufacturers have time and resources to gradually work out the kinks and improve design almost to perfection. With computers, on the other hand, "physical" amortization is low, so the only way to sell new systems is to cram them with more new features, no matter how poorly designed at first, and to make existing ones (no matter how proven and reliable) obsolete and incompatible. Simply reducing the number of bugs will not generate many sales. As a new feature appears, buggy and frustrating to use at first, the economic machine of the computer industry kicks into high gear. Magazines write raving reviews to increase their own sales, add-on manufacturers rush to incorporate it and propagate it down the sales channels, application developers write new drivers and other utilities which make new feature indispensable and previous versions obsolete. As a result, today complex software is not unlike a human genome - a product of often messy and chaotic evolution, rather than a compact and elegant design. Pieces of active, useful code ("genes") are surrounded by "junk", leftover from previous generations of development, often redundant and useless. Why it is there? Because it is easier and cheaper to throw more hardware to crunch ever-bloating volumes of code and not to touch old rusting scrap, than to design and debug fast, efficient code. And it is not getting any better. On the other hand, despite all these intrinsic problems and flaws, many complaints against computers are quite unreasonable. For example, the story often goes, it is difficult to find that text file created two months ago, or where are those digital pictures from the last trip. But this supposedly unfavorable comparison with the "real world" does not hold. Consider, for example, the tree-like directory/subdirectory/file hierarchy, used in most operating systems. In fact it closely resembles a real-world storage system - file cabinet/shelf/file/document, only better. Why are we complaining? Because we have much higher expectations of computerized data storage, than of a traditional file cabinet. A file cabinet requires careful maintenance; if we treat it the same casual way we do computer files, it would be totally unusable in two weeks. Complaints against computers notwithstanding, it is far easier to find past notes and other files on a computer than in a "physical" world. The same with the gripes against Internet search engines, repeated in the book - a familiar story about a list of 10.000 irrelevant links in response to a search query. I think it is just a trite cliché. Frankly, it never failed for me to quickly find stuff even without following "exact" grammar rules recommended by engines. Besides, there is a good chance to discover surprises, interesting and useful information among those "10.000 links". Of course, one could have a negative experience with web searches. In the "real world" a stupid or badly posed question is unlikely to produce a useful answer. Why do we expect a different result from a search engine? Moreover, search engines in the last few years was among the most competitive and dynamic technologies, where leaders changed almost every year - Yahoo, Altavista, HotBot, Northern Light, Google, each progressively offering better, faster, more complete results. The author touts XML and "semantic web" technologies as one of the "saviors" to untangle the computer industry mess. Again, I have serious doubts about this proposed magic bullet. The beauty of the first versions of HTML, when it appeared in early 90's, was its simplicity and universality. Any intelligent person could master it in half a day, and publish a decent-looking web page, which could be seen on PC, Mac or UNIX workstations anywhere in the world. This was truly revolutionary. The XML and "semantic web" at the first glance is just a natural extension along this road. But instead introduces another big layer of complexity, reduces the pool of programmers who can quickly master it, opens the door to innumerable new bugs and inefficiencies. If HTML opened a new chapter in computer history, XML and its companion technologies do not. It is filling the same chapter with comments and footnotes until the text becomes illegible.
|
|
|
8 of 9 people found the following review helpful:
4.0 out of 5 stars
Great Thoughts, Limited Reality, More to Do...., January 14, 2002
In some ways this is the gold-collared knowledge worker counterpart book to Ted Halstead and Michael Lind's The Radical Center: The Future of American Politics (citizen-centered). Those who liked The Cultural Creatives or IMAGINE: What America Could be in the 21st Century, can adopt this book as their user's guide for demanding change in information technology.
I recommend it because it is full of common sense, is the first really helpful "requirements document" for a clean sheet new approach to software and hardware and ergonomics ($3000 word for user friendly). The bad news is that nobody is listening. We are ten years away from this being a reality because the legacy providers (big hardware, one certain software company) are not about to retool their empires for the sake of delivering better value. It is more than a little amusing to me to have this book endorsed by the CEO of the one company that prides itself on producing software with mutated migrated Application Program Interfaces that are used to extort tribute from third party software developers, where no sane consumer will invest in his products until they've had three years to "mature" in the marketplace. The opening listings of the "standard faults" in today's "consumer electronics" is alone worth the price of the book--unintegrated systems fault; manual labor fault; human servitude fault; crash fault; excessive learning fault; feature overload fault; fake intelligence fault; waiting fault; ratchet fault... The book ends on a low note and high note. The low note is a description of Oxygen, a $50M project seeded by DARPA and including several major company partners such as HP and Nokia. This project has some excellent ideas, including a new focus on an architecture for nomadic computing with three aspects: a Handy 21 (hand-held), Enviro 21 (intermediate personal computers at home, office, and in car), and N21 Network (Intentional Naming System, every computer and peripheral everywhere is in the public domain and broadcasting its location and status, use on the fly). Good stuff. What he doesn't mention is that the U.S. Government is spending over half a billion dollars on completely uncoordinated desktop analysis toolkits, and there is probably 2-3X that much being spent in the private sector. He does note that we will never get our act together if we continue to develop hardware and software in a very fragmented and hardware-based manner. On the high note, the author has clearly thought about the consequences of having an information revolution here in the USA, creating information royalty, while leaving the rest of the world dispossessed, in poverty, and unconnected. He has a very practical appreciation for the fact that the USA must fund two distinct foreign assistance programs--a Digital Marshall Plan (my phrase) to jack in the entire world; and a commensurate literacy, birth control, disease control, and famine control program to stabilize populations to the point where they can be productive within the global grid. I read this book on the airplane coming back from the Consumer Electronics Show in Las Vegas (Federal Emerging Technologies Conference sub-set), and I was really struck by the contradiction between the vast fragmentation spread out over Las Vegas (the man who has everything also has to carry it) and the elegant simplicity of this book's vision--one hand-held able to be any of 100+ devices. "It's the software, simpleton...." What saddens me, especially when considering the billions of dollars being given away by our richest software developer, someone who seems to favor gestures on the margin instead of quality control and open source at the core, is that we knew all this in the mid-1980's. The eighteen distinct functionalities needed for a desktop analysts' workstation were identified by CIA in 1986--everything from data ingestion and conversion softwares to modeling and simulation and pattern detection and of course desktop publishing. The year after the CIA prototypes were working so successfully on UNIX (Sun), CIA decided that the PS2 would be the standard "dumb" terminal, and all UNIX efforts were ordered to shut-down. The big organizations, the ones with the power to make the revolution, chose control and dumb terminals over freedom and smart software. I am very skeptical that the vision in this book will come to fruition...
|
|
|
Most Recent Customer Reviews
|
by Jeanne M. Dams
by Bill Martin
See all Editorial Reviews
