Analysis of Results - Java: One Year Out | WebReference

Analysis of Results - Java: One Year Out

The Measurements

This study used both a quantitative measure of Java's diffusion for indications of adoption among its social system and a qualitative measure comparing Java against the diffusion theory model.

Quantitative Measures

The quantitative measures used were:

  • The total number of applets listed in the Gamelan Directory gathered on a bi-weekly basis from February 1, 1996 to April 15, 1996 with an additional reference point taken on July 15, 1996.
  • The number of applets listed in individual Gamelan categories from February 1, 1996 to April 15, 1996 with an additional reference point taken on July 15, 1996.
  • The number of posts per week to the Usenet newsgroup, compared to the number of posts per week to the comp.lang.c++.
  • The total number of Java usergroups listed on the Javasoft Web site.

These quantitative measures were used to gather data about the rate and number of applets being created, the use of these applets, how much discussion Java generated among its potential adopters, the number of formal groups constructed to focus on Java, and the number of jobs Java technology initiated.

Quantitative Results: EarthWeb's Gamelan Directory

Table 1 shows dramatic growth for EarthWeb's Gamelan Java Directory over a six-week period. The directory grew by a total 1291 applets and contained 2372 resources on April 15, 1996. This is an average addition of 170 applets every two weeks. The data acquired on July 15, 1996 shows continued growth with an addition of 1340 applets. Growth has slowed somewhat in this period, as nearly the same amount of resources were added in a twelve-week timespan that were added in a six-week timespan. However, the numbers have not leveled off and this table suggests a fast adoption rate and an ultimately successful innovation.

Table 1

Table 2, detailing resources in individual Gamelan categories, reveals more insight into how Java is currently being adopted.

Table 2 - Java Resources vs. Category in Gamelan

Category Number of Entries
2/1/96 2/15/96 3/1/96 3/15/96 4/1/96 4/15/96 7/15/96
Arts & Entertainment 128 140 170 184 206 305 532
Business & Finance 21 23 27 28 34 39 67
Educational 90 105 115 128 147 209 337
JavaScript 27 46 54 62 84 135 242
Miscellaneous 45 50 65 67 73 82 77
Multimedia 117 137 155 170 190 260 360
Networks & Communication 29 42 36 40 52 96 177
News 4 5 43 45 47 52 79
Programming in Java 194 216 243 252 283 343 491
Special Effects 136 158 145 159 182 212 292
Utilities 66 73 71 79 99 140 256
Web Sites 224 251 277 290 340 499 766
Total Resources 1081 1246 1401 1504 1737 2372 3676

In Table 2, the Web Site category is both the largest total category and the largest growing category. These figures represent one of the most general Gamelan categories, and as such, this category is a fairly useless indicator of Java's current applications.

The Programming in Java category remains one of the highest total categories in both data set examples. The nature of this category indicates that some serious attention is being paid to learning the language and documenting that knowledge for other developers. This group has the most to gain from using and contributing to the Programming in Java category. Because it is one of the fastest growing Gamelan categories, Java application development looks promising. However, this category is only a positive indicator of Java's adoption rate if its growth does not result from an exchange of development information because the language is too complex.

In both data sets (the first ending on ending on April 15, 1996 and the second on July 15, 1996), Arts & Entertainment, Multimedia, and Special Effects were among the top resources. These category totals, although indicative of adoption, do not point to application development which demonstrates the power of transparency built into the language. However, the second data set shows important resources such as Networks & Communication and Utilities growing at a faster rate. Applications which utilize those features will further the innovation's acceptance.

Graph 1 & comp.lang.c++ is the Usenet newsgroup hosting discussions about the Java language and its developments. The figures taken from were used to measure the amount of discussion Java generated among potential adopters. For comparison, seven day averages were calculated for the comp.lang.c++ newsgroup. The Java programming language, with its roots in C++ and object-oriented programming, is compared widely to C++ by its creators and users alike. As such, the focus and members of comp.lang.c++ newsgroup closely match those of the newsgroup. These measures match the current prominent programming language against the emerging programming language to determine how much the amount of discussion compares against the industry standard.

Quantitative Results: & comp.lang.c++

On average, the daily number of posts to the newsgroup outnumber the posts to the comp.lang.c++ newsgroup by 98 per day. The numbers for both groups remain fairly flat over the first seven-week period. The second data set shows a decline in the number of posts to the Java newsgroup. However, what the data does not show is that in this time period eight additional newsgroups in the hierarchy have been added.

Graph 2

These numbers illustrate the hypothesis that Java is one of the most discussed developments.

Java Usergroups

The number of Java usergroups was tallied at the beginning and end of the study, in order to measure any increase. Usergroups are comprised of individuals from various programming and professional backgrounds with an interest in learning the Java programming language. These usergroups contain, therefore, potential adopters and members of the social system.

Quantitative Results: Java Usergroups

While these numbers and their respective increases may appear small, the very fact of their existence indicates high interest in Java. These usergroups consist of a large network of people and organizations. Typically, members are professionally involved with Java and/or programming. They share knowledge with members of the group and also serve as gatekeepers in distributing information to those not of the group. Many of these usergroups maintain Web pages which then furthers the spread of information. These usergroups are a slightly more formal form of interpersonal communication in that they generally have regular meetings and topics of discussion. The formation of one usergroup around an innovation indicates interest and commitment. The Java innovation showed an increase in the number of usergroups across the world.

Table 3 - Java User groups listed on Javasoft

2/5/96 4/9/96 7/15/96
Usergroups listed on Javasoft
  1. Australia
  2. Hawaii
  3. Toronto
  4. NYC
  5. Staten Island
  6. Mid-Atlantic
  7. Baltimore/Annapolis
  8. North Carolina- RTP
  9. Cincinnati
  10. Dallas
  11. Missouri & Kansas- Midwest Regional
  12. Oregon
  13. Sacramento
  14. San Francisco
  15. Mexico City - Monterrey Institute of Technology
  16. London
  17. University of the Basque Country
  1. Sun User Group's Special Interest Group
  2. Australia
  3. Hawaii
  4. Toronto
  5. NYC
  6. Staten Island
  7. Mid-Atlantic
  8. Southern Virginia
  9. Baltimore/Annapolis
  10. North Carolina
  11. Cincinnati
  12. Cleveland
  13. Atlanta Area
  14. Chicago
  15. Missouri & Kansas- Midwest Regional
  16. Dallas
  17. Oregon
  18. Sacramento
  19. San Francisco
  20. Mountain View/Silicon Valley
  21. Mexico City - Monterrey Institute of Technology
  22. London
  23. University of the Basque Country
  1. Sun User Group's Special Interest Group
  2. Australia - AJUG
  3. Hawaii
  4. Toronto
  5. Vancouver
  6. Winnipeg
  7. Ottawa
  8. Boston
  9. Connecticut
  10. NYC
  11. Staten Island
  12. Philadelphia Area
  13. Mid-Atlantic
  14. Southern Virginia
  15. Baltimore/Annapolis
  16. North Carolina - RTP
  17. Tampa Bay
  18. Minneapolis-St Paul
  19. Wisconsin
  20. Illinois
  21. Chicago Great Lakes
  22. Cincinnati
  23. Cleveland
  24. Atlanta Area
  25. Missouri & Kansas- Midwest Regional
  26. Dallas
  27. Anchorage
  28. Oregon
  29. Denver Area
  30. Sacramento
  31. San Francisco
  32. Mountain View/Silicon Valley
  33. San Diego
  34. Mexico City - Monterrey Institute of Technology
  35. London, England
  36. University of the Basque Country

Qualitative Measures

The remainder of the study used Rogers' diffusion theory to define the innovation and project its future adoption based on the way in which it corresponds to successful innovations, its reception in the social system, the members and communication channels of the social system, and finally how time is a factor in the innovation's diffusion. Successful innovations are perceived to have a high degree of relative advantage, compatability, trialability, and observability; and a low degree of complexity.

The Innovation: Java & Relative Advantage

Relative advantage: "the degree to which an innovation is perceived as being better than the idea it supersedes." [14]

Java is a solid heavyweight in the relative advantage category.

During Java's diffusion process, reports concerning its perceived relative advantage appeared in stages. After its formal announcement in May 1995, Java was first perceived as the browser to bring interactivity to Web pages by a social system familiar with the browser concept. This part of the innovation was immediately viewed as new and better than other Web browsers because it promoted user interactivity.

Eric Schmidt, chief technology officer of Sun Microsystems, has seen the future of the World Wide Web --and he thinks it's hot. HotJava that is -- the much-discussed interactive Web browser Sun unveils officially this week in a burst of eleven Internet announcements that include turnkey Web servers and security firewalls. . . . HotJava taps the computing power of a Web surfer's machine to allow true interactivity -- animation or playing live games against others -- on the Web, all things that move beyond static text and photos.[15]

As the diffusion spread, features of the programming language (simple, object-oriented, portable, robust, architecture neutral, multi-threaded, dynamic, and secure) encompassed the second stage in Java's perceived advantages. Java's White Paper explicitly itemizes Java characteristics which make it better than the C++ language. In this, Sun positions Java as an innovation which supersedes a previous technology.

Imagine you're a software application developer. Your programming language of choice (or the language that's been foisted on you) is C or C++ . . . . The tools you use to develop applications don't seem to help you much. You're still coping with the same old problems; the fashionable new object-oriented techniques seem to have added new problems without solving the old ones. You say to yourself and your friends, "There has to be a better way!" Now there is a better way--it's the Java programming language environment ("Java" for short) from Sun Microsystems.[16]

Java's conception as a tool derived specifically for distributed environments added yet another perceived advantage. Although its initial incarnation as the underlying architecture for consumer appliances failed to achieve mainstream status, this architecture was ideal for its emergence on the Web. Its originality was also suited for the new innovation that was the Web. In particular, potential adopters cited Java's architecture neutrality, portability, and security features as advantages. Java quickly advanced from the latest Web buzz to the perception that Java was its most useful tool. Those three features promoted new ideas less in the direction of animated graphics and more in the direction of applets that could be executed on any machine, across networks, because of the platform independence and security.

Using Sun's Java technology, users could access programs resident on server computers. Part of the software would run on the user's machine. When the program was finished, it would simply vanish. That would reduce the need for hard disks and floppy drives. Developers are already producing Java applets that people can download and run on their computers. Sun would apparently provide the Java end of things and team up with a PC maker to produce the hardware.[17] In addition, a large number of companies continue to license the Java technology for future or current projects, including, Netscape, Microsoft, Novell, Oracle, Symantec, Borland, Adobe, IBM, and Macromedia. [18] These companies perceive an advantageous benefit in using the Java technology or at least retaining the ability to use the Java technology to remain competitive.

Microsoft will announce by the end of the month a plan to embed Java directly into future versions of Windows, a significant development that will allow other applications besides Web browsers to run applets downloaded from the Internet. The Microsoft announcement is part of a major strategic shift for Sun Microsystems, which will make a concerted push with the Redmond, Washington, company and other operating system vendors at Internet World in San Jose in two weeks to make Java a standard feature of operating systems rather than Web browsers. [19]

At the present stage in the diffusion process, the language is used more as an abstract concept or vision rather than a concrete tool. Java's perceived advantage is driving the industry to develop new concepts of computing and to dust off old concepts (such as interactive television and personal digital assistants) that might gain acceptance with an underlying architecture that has already amassed industry appeal. The language's concepts are being explored much more visibly than actual Java applications. Java has developed a model of computing that shifts the focus of computing to a heterogeneous networked environment, beyond the current local area and wide area networks to a truly transparent client/server model. The implications of network computing, the Network Computer, the browser operating system, the merging of PCs and television, and information appliances such as PDAs are being explored because Java technology is perceived as being better than the idea it supersedes.

The Innovation: Java & Compatibility

Compatibility: "the degree to which an innovation is perceived as being consistent with existing values, past experiences, and needs of potential adopters." [20]

Java is a solid heavyweight in the compatibility category.

As a programming language, Java is compatible with existing values in the object-oriented environment. The innovation is compatible with a large percentage of past experiences of programmers and application developers in its similarity to C++. Sun certainly stresses this compatability in the Java White Paper.

Another major design goal is that Java look familiar to a majority of programmers in the personal computer and workstation arenas, where a large fraction of system programmers and application programmers are familiar with C and C++. Thus, Java "looks like" C++. [21]

Java's diffusion is also compatible with the open standards of the Internet community. Since the Internet's inception, those involved valued the free exchange of knowledge and tools. This exchange created the Web. The free distribution of the Mosaic innovation added millions of members to the Web social system. Java's release follows in this tradition, as Sun made the JDK and its source and binary codes available to all.

Continuing and constant innovation is a value held by the social system. Because of the Web's relatively recent creation, being "new" is foremost among the social system's past experiences. As an innovation diffused on the Web, Java is compatible with the belief in pushing the boundaries of the medium to realize new possibilities.

The Innovation: Java & Complexity

Complexity: "the degree to which an innovation is perceived as difficult to understand and use." [22]

Java is a bit of a lightweight in the complexity category.

The first buzzword the Java White Paper uses to describe the language is "simple." Clearly Sun aimed to position Java for widespread adoption. The language may be simple for a C++ programmer, but for beginning programmers and those without object-oriented backgrounds, the language is not a simple one to learn, understand, or use.

But can designers learn Java? Only if they're programmers! Java is a *real* programming language, and what's more it's object-oriented. This is not simple stuff, to do it well, one needs some background training. [23]

In order to reduce the complexity, Sun provided in-depth documentation on the language's features, implementations, class libraries, and API. [24] Sun makes the source code to all of its applets freely available for others to peruse, copy, or modify. Following Sun's example, the majority of individuals creating applets provide access to their source code as well.

As testament to Java's popularity, very soon after Java's release, numerous "how-to" books were published and on-line tutorials developed. As of July 1996, 43 Java programming books were in print with 71 more expected in the next twelve months. In April, when this data point was first taken, there were only 17 Java books in print. These books range in complexity, targeted to audiences ranging from HTML authors and designers to more advanced object-oriented programmers.

Instant Java is written primarily for HTML authors who want to include Java applets on their Web pages. I have tried to create useful, general purpose applets that will appeal to the majority of Web authors. [25]

Visual development tools will move Java closer to fourth-generation software languages which conceal most of the complexity from the user while still allowing for powerful application development. They also have the potential for including the non-professional programmer among the ranks of Java developers. For example, Symantec found that a large number of Cafe's customers included designers and developers with very little C++ experience. Their original Java development environment, shipped as an add-on to their existing C++ product, was redesigned with a focus on the needs of the more inexperienced developer.

To be honest, we did not understand who would be using Java. We simply assumed that professional C++ developers would be moving to Java. We quickly got very strong feedback that people were totally confused. We were hearing from MIS and Visual Basic developers and even HTML page designers that had no practical C++ experience. [26]

As yet, these tools are in their infancy and need time to mature and stabilize. But in order for Java to overcome its complexity and achieve widespread usage, these development tools must become premiere products.

While the language is not the simple affair first suggested by Sun's buzzword, enough resources exist to reduce complexity and promote use. But will they? Out of the five components for a successful innovation, complexity is the most important one for Java. Regardless of perceived advantages, if Java is too difficult to learn, it will not gain widespread use and will ultimately fail.

The Innovation: Java & Trialability

Trialability: "the degree to which an innovation may be experimented with on a limited basis." [27]

Java is a solid heavyweight in the trialability category.

Sun built a number of features into Java's release that not only permit but encourage widespread experimentation. For example, the javap (a tool which decodes compiled Java code) allows developers to examine the code of any applet or application and use these implementations as a reference. Sun also set the example for the social system by providing the source code for its applets for others to copy, modify, or use. It is now a common practice for other developers to include the source code for Java applets as reference.

In order to allow educational and nonprofit institutions to learn how the system works and for those interested in developing JDK ports to other systems, Sun released the JDK source freely. One result of this move is that people are enabled to experiment with the language and learn its strengthes and weaknesses. By finding holes, such as that found by the Princeton team [28], the language can become more robust for future implementations. Commercial developers can license the Java source code for a fee.

By opening Java development to all interested parties, Java stands to become the standard architecture for network-centric applications. Potential adopters are able to experiment with the language and derive their own conclusions about the innovation's usefulness.

The Innovation: Java & Observability

Observability: "the degree to which the results of an innovation are visible to others." [29]

Java is a solid heavyweight in the observability category.

Java applets are visible to all Web users using the Netscape Navigator 2.n Web browser. Netscape announced its plans to license Java immediately following Java's release in May 1995, and in December 1995 released their Navigator 2.0 browser, capable of viewing beta Java applets. Netscape's commanding share of the browser business indicates that a large majority of Web users can view the results of Java technology. Currently, Java applets cannot be displayed on machines running Windows 3.1 even when using Navigator 2.0. With Microsoft's Internet Explorer 3.02 beta release, millions of users with Netscape and Explorer browsers have the opportunity to observe Java in action.

There are a number of Java archives, such as Gamelan, where users can go to view Java applets. Very often, a user will inadvertently land on a Web site containing Java applets and have an opportunity to view the results of the innovation. Regardless of whether the user seeks out an opportunity for evaluation or whether one coincidentally presents itself, Java's results are visible to a large population throughout the Web's user base.

Not all visibility is advantageous to the innovation nor desired by the user. Java is in a formative period as developers learn the language. Some users find implementations of the language inconsequential and even annoying at best. A user not seeking out that applet can form a negative impression of the innovation because it impedes the task at hand.

Application development off the Web is not as visible to potential adopters. Industry leaders publicize their plans for Java development, however; very few results are visible.

Communication Channels

For a successful innovation, information must be easily available to all potential adopters. In Java's case, these communication channels are staggering in their diversity and in their ability to instantaneously spread information to millions of potential adopters. Traditionally, information on an innovation has spread through interpersonal channels and only reaches the mass media once a sizable number of adopters warrants the media's attention. The communication channels involved in this diffusion process contain three different mass media channels: mainstream, trade, and Web sites; and at least 5 interpersonal channels: newsgroups, listservs, personal Web sites, usergroups, and traditional face-to-face communication.

In the year since Java's release, 50 newspaper articles reporting specifically on the programming language have appeared in mainstream newspapers, including The Wall Street Journal, The New York Times, USA Today, Los Angeles Times, Boston Globe, and Chicago Tribune. Because this population includes people entirely uninvolved with the Web and the computing industry, Java's diffusion and its social system contain the potential to increase.

Industry trade publications and other media offer more in-depth insight into the innovation and often engage in future projections. From May 1995 to March 1996, 218 articles with Java as the main topic appeared in computer periodicals. An additional 970 articles made mention of or discussed Java within the scope of other related information.

Web sites can constitute mass media channels in their ability to convey information to a wide audience in a short amount of time. Certain types of Web sites are more akin to the typical mass media channels in content and scope whereas other Web sites fall more along the interpersonal communication lines. The Javasoft site, for example, documents its innovation, offers on-line tutorials, publicizes the company, and announces new developments. Established Web sites (such as Yahoo or Gamelan) serve a similar purpose. Users know that these sites will contain desired information and thereby seek them out on a regular basis.

Information about the innovation is also being carried through interpersonal channels in the form of newsgroups, listservs, individual Web sites, and usergroups. The majority of this communication is in electronic form resulting in a great expansion of the traditional face-to-face interpersonal communication channels. The number of interpersonal channels also includes those informal face-to-face communications among professional and student colleagues. Such a wide information network increases the amount of information about the innovation, exploration into the innovation, and the amount of time necessary for the dispersal of that information. These communication channels are a major factor in the Java hype that has guaranteed momentum.

Social System

Java's social system is made up of all Web users (users, designers, and developers) regardless of their level of experience or technical knowledge. This user base constitutes the potential adopters ultimately accepting or rejecting the innovation.

Within this system varying relationships and statuses exist. The majority of the user base contains average Web users receiving information about the innovation coincidentally through exploration of the Web. These users, although potential adopters, are more casual bystanders of the Java innovation. Because of the coincidence of their knowledge, this user base is less likely to influence ultimate adoption. This user base is important though in an environment where hype equals momentum. Communication maintains a collective conciousness about the innovation within the social system. This group will most likely follow the opinion leaders' decisions to adopt or reject.

Another set of average users actively seeks out information through Web sites, media channels, and interpersonal communication. This subset of potential adopters has not yet formed a decision to either adopt or reject the innovation.

I am trying to learn my way around www and internet and computer and bill gates, etc. What is JAVA. Do i need JAVA. Is it something already in my computer? [30]

Some members of the social system seek out information and try to learn the programming language or use already developed applets on their Web pages. This set of potential adopters have made a initial decision to adopt on a limited basis. Other members of the social system are experienced developers who have received enough information to discuss the various reasons for or against adoption of the innovation. These members actively engage in trading viewpoints throughout the diffusion process. As the group most likely to develop products using the innovation, the decisions made by this set of potential adopters is critical to the success of the diffusion process.

>Portability doesn't matter.
As a developer of a cross platform product, what matters to me is industrial strength portable languages, libraries, and so on.[31]

Opinion leaders and change-agents also occupy an influential sphere in the diffusion process. Examples from these sets would be the major companies involved in the diffusion process, Sun Microsystems, Inc., Netscape Communications Corporation, and Microsoft, Inc. These change-agents are controlling the diffusion rate with actual products, support for products, and announcements regarding future products. As opinion leaders, their views on the industry influence the communication channels either directly, such as Bill Gates' weekly column, "Ask Bill" in The New York Times, or indirectly through press releases, product announcements, and company alliances. These events are discussed and interpreted both in mass media channels and interpersonal channels. They influence potential adopters' decisions about the innovation.


This study used the time factor to measure the current rate of adoption as a reference point in Java's diffusion process. Java is sixteen months into this process. In that time, a number of innovators have adopted and further promoted the innovation, e.g., Netscape Communications Corporation, Symantec, Oracle, and Borland. All those who have developed professional Java applications which enhance or add to the innovation are innovators. Presently, at sixteen months out, the diffusion process is yet in its early adopter phase. Those who have developed Java applets or applications whether for personal or professional use, are considered early adopters.

The current rate of adoption indicates that the majority of adopters are using Java for entertainment, special effects, and multimedia purposes. These developments are mainly demonstration applets on individual Web pages. Industry documentation reveals high interest in using Java for industrial strength application development; yet such tools are in early phases consisting mainly of announcements. In this first year, crucial innovation support systems are being developed, such as:

  • an expectant user base
  • visual development tools
  • improvement in Just-In-Time (JIT) compilers
  • better browsers
  • committment to open standards

In a fast-paced environment, relative newness does not last very long, giving early adoption a small window of opportunity. Java must be able to capitalize on its window so that it is the innovation used as the underlying architecture for network-centric computing devices rather than the innovation that merely shifted the industry's focus.

Comments are welcome

Copyright © 1996 Bridget W. Regan and
Created: May 17, 1996
Revised: July 31, 1996