Issue
Article
Vol.28 No.3, July 1996
Article
Issue
Issue |
Article |
Vol.28 No.3, July 1996 |
Article |
Issue |
`Multimedia and Hypertext, the Internet and Beyond' is a book about hypertext usability: the usability of the software as well as the hypertext itself. The Internet (read: the World Wide Web) is only one of the examples. The `beyond' in the title refers to a chapter with predictions about the growth of hypertext usage.
The book starts with an attempt to define hypertext. Obviously it is not possible to give a complete definition, but the author succeeds in listing a number of useful characteristics. At the end of that section he states that the best test is whether it `looks and feels' like hypertext. It may seem a weak subterfuge, but it makes it perfectly clear that `hypertext-ness' is not in the system architecture, but in the user interface. An important point, since many people on the World Wide Web think that everything that has a URL is a hyperlink.
The second chapter introduces hypertext from the viewpoint of the hypertext author, using a small text by the author himself as a case study.
The third chapter outlines the history of hypertext, starting with Vannevar Bush in 1945. To people reading this review that history is probably well known, but Nielsen's audience is wider and thus it makes sense to include the chapter. (Although he never defines his audience, the flashy title of the book must have been chosen for a reason.)
Chapters 4 through 12 deal with the creation of hypertext. It tells about when not to use it, what hardware and software one needs, what the user interface problems are, how one transforms existing information to hypertext, etc. There are lots of examples of different applications, from on-line help to cooperative authoring and even games. The many examples make the text easy to follow and indeed are a resource in themselves.
Right in the middle is a chapter on `Hypertext on the Internet' (chapter 7). In the sequence of chapters this one looks a bit out of place. Where other chapters describe issues common to most, if not all hypertext, this one concentrates on a single system, the World Wide Web (with a few words about Hyper-G).
However, on closer reading, it does mention some new problems that are of more general applicability: paying for information you actually use, when paying for the total is infeasible; scalability, in amount of information as well as in number of concurrent users; preparing for different media; and support for different languages, in the software as well as in the hypertext itself.
In the section on scalability (Flash Crowds, page 175) Nielsen introduces a new term that may very well become a buzz word. A `flash crowd' is a sudden run on a certain Internet resource. The author defines `sudden' as during a day or two. In reality flash crowds can happen within seconds. (The author attributes the words `flash crowd' to a SciFi story about teleportation by Larry Niven, but it is fair to credit Nielsen himself with applying it first to virtual crowds.)
By the way, this chapter also shows a disadvantage of paper books compared to hypertext that the author forgot to mention in his preface: the impossibility of correcting errors or updating information that has become out of date. But this chapter also proves that a good author can write a useful analysis even if the facts change while he's writing it.
The last three lines of page 199 have probably been added as the very last thing before the book went to press. They show the only URL with the new `www.w3.org' instead of `info.cern.ch' and they mention a new development that the author apparently recognized as important but couldn't integrate into the text anymore (style sheets).
The same chapter also introduces the concept of `half-dead links'. A half-dead link is something in between a traditional passive reference and a hyperlink. It is not a hyperlink, since the linked information does not appear immediately on activation (as chapter 1 required). But it is not completely inactive either, since the system will fetch the target document eventually, asynchronously and usually after some intervening actions, maybe within minutes, maybe the next day.
Chapter 13 deals with the future. It tries to predict the amount of `market penetration' of hypertext over the next 20 years. Most predictions seem pretty straightforward. For example, electronic payments are still in early experimental stages, but it is reasonable to expect working payment systems within five years. (That is four years to go, since the book is a year old.)
Another prediction is that hypertext will be integrated with other applications. Before 1995 there was not much evidence of that, but in the year since the book's publication a lot has happened to support the prediction. The Web is replacing the built-in help systems of many programs. Conversely, programs are being run within Web browsers, so that hypertext navigation takes the place of entering commands or command arguments. Sun's Java programming language, for example, has accelerated that process enormously.
Actually, the current integration already goes beyond the cautious prediction, which talks only of merging between hypertext and certain types of advanced applications (AI, simulation). It is rather the common, everyday applications (mail, news, word-processing) that are combined with hypertext, thus benefiting many more users.
Whether the prediction of major changes to the system of copyright and related rights will come true in the next 10 years is much less clear. The author's reasoning is logical enough, `information wants to be free', but there are important forces that have a strong interest in keeping information prisoner. They may succeed in crippling electronic information well beyond the author's estimate of the year 2000.
The book ends with almost ninety pages of annotated bibliography including many on-line resources. The book also has a good index.
Who should read this book? First of all, I would recommend it to everybody developing software for the Web. And they should read the whole book, not just chapter 7. Researchers tend to read the literature before building something new, but Web developers seem to think the Web is the first, if not the only example of hypertext in existence.
The book doesn't contain compact lists of guidelines, like some of Ben Shneiderman's texts do, so it is probably less useful for the average hypertext author, unless they are interested enough to read the whole book. On the other hand, if you are creating a large hypertext, on the Web or elsewhere, then it is certainly a good idea to read the whole thing.
Researchers in usability and hypertext will have to set aside any reservations they have because of the book's title. This is a useful book to have in your personal library. I have already had occasion to refer to it several times.
The publishers have evidently tried to keep the book inexpensive. That is a laudable goal, but it is unfortunate that they didn't use a slightly better DTP package. A book of lasting value (as I think this one is), and moreover one dealing with user interfaces, should have had more attention paid to the layout. As the author himself says (page 263): "Printed books look different depending on their quality and age."
Jakob Nielsen `Multimedia and hypertext, the Internet and Academic Press Professional, Boston, 1995, ISBN 0-12-518408-5, 480pp, NLG 67,45
About the Reviewer
Bert Bos received a Ph.D. in 1993 from the Faculty of Arts of the University of Groningen, on the subject: `Rapid user interface development with the script language Gist'. After that he worked on a project to develop software and courses for helping scholars in the Humanities make use of the Web, before joining the World Wide Web Consortium (W3C) in 1995. He works for W3C at INRIA in France and is in charge of the Internationalization area.
Email: bert@w3.org .
Issue |
Article |
Vol.28 No.3, July 1996 |
Article |
Issue |