1. Managing Through Cycles Of Technological Change, Anderson and Tushman, Research Technology Management, 1991 (Online)

1. Profiting from Technological Innovation, Teece, Research Policy, 1986 (coursepack)

• This is a demanding article, so plan your study time accordingly. You can skip Section 6 on policy implications.

1. Case Study: Pixar, Inc (handed out)

Study Questions for Andersen and Tushman

1. What constitutes a "technological discontinuity"? What constitutes a "dominant design"? What characterizes the nature of technological progress in the "eras of ferment" and "eras of incremental change" demarcated by technological discontinuities and the emergence of dominant designs?

• As a back drop to the questions below we will focus on the case of VA Linux, the pioneering company building hardware systems based on the Linux operating system (see attached information on open source and VA Linux)

1. What are "appropriability regimes"? What leads to strong versus weak appropriability? What are some current examples of technologies with tight appropriability regimes today? Weak regimes? Why? How would you characterize the "appropriability regime" for VA Linux?
2. What are complementary assets? What distinguishes generic, specialized and cospecialized assets? Why does it matter what kind of complementary asset an asset is? What are the most important complementary assets for a systems vendor like VA Linux, and what kind are they?
3. How does the basis of competition shift before and after the emergence of a dominant paradigm/design? How does this relate to how the VA Linux story unfolded?

2. It's 1996 and Pixar's stock price has been languishing in the teens, 50% of its IPO day high. The market appears to be concerned about Pixar's ability to fund and sustain the technical and creative development required to create full-length digital animated features going forward. What is the best strategy for Pixar in 1996? What lines of business should they be in?
3. Was Disney the right partner for Pixar in 1991? Is Disney the right partner for Pixar at the end of the case in 1996? What about Disney? Should they continue to partner with Pixar or pursue some other strategy with regard to computer generated animation?

• Write a two page, single spaced (approximately 1000 words) essay addressing questions 1, 2 and 3 above related to the Pixar Case. You are welcome to make use of outlines, bullets, tables or any other device to get your points across quickly and easily. Just make sure that I'll be able to understand what you're saying.

One of the more important developments in the IT field in the last decade is the emergence of the so-called "open source" model of commercial software development. Some have called "open source" a disruptive technology, which may or may not be right, but in any case it is an interesting development and provides the backdrop for our mini-case example, systems vendor VA Linux.

With the conventional software development model, a company, such as Microsoft, builds a software package, such as the Windows operating system, and takes on all responsibility for future enhancements and improvements to that package. While Microsoft will work closely with industry partners throughout the development process, they are in total control and they of course own the result of the effort, for which they can charge as a they see fit. While companies such as Microsoft will publish information about interfaces to the system, the actual program source code itself will not be released. Rather, the code itself is treated as a trade secret. Commercial software vendors will also get copyright protection for their systems, which prevents another company from independently developing a package that is substantially the same in terms of "look and feel".

Under the open source model, by contrast, an single programmer or a small group of programmers will develop a software system and make that package freely available to the public. Not only will compiled code be made freely available, but the actual program "source" code itself will be made available as well. What this means is that users can figure out for themselves how a program works and whether it is working correctly. They can modify the program at will to fix bugs or to better suit their needs. There will still be a "standard" version of the system for which modifications and improvements are administered either by the original developers or by a group they have set up for this purpose. For the most popular open source packages, such as the Linux operating system and Apache web server, there has emerged a worldwide army of programmers that are constantly trying to find errors or suggest improvements. It is considered a high honor among these folks to have their contributions incorporated into the next version of Linux.

Commercial companies also play a role in the open source model. While in principle any company can download a version of Linux or Apache from a free site, this imposes upon them the burden of installing and configuring the software and complementary products (such as utility programs or a graphical user interfaces). Since many end users would prefer not to do this, open source software vendors such as Ret Hat have emerged that take on the burden of "packaging up" the open source system. They will develop their own version or "build" of the software that includes extra programs to make installation of the software itself and complementary tools go easier. And they will offer after-the-sale service and support. They must charge for all this, of course, but what they are charging for is not for the open source software per se, but for their particular configuration of the software. And buyers are still free to further modify the program as they see fit.

Critics of the open source movement will say that it is only appropriate for certain kinds of software (e.g., systems software and utilities that do not have to respond quickly to the needs of the market), that most IT managers would never trust their core applications to software than no particular entity controls and can be held accountable for, and that the coordination effort to keep these systems up with the times will prove too great.

In any case, there is no disputing popularity of the Linux operating system, the best known and most important open source software system. Developed in the 1980s by a Linus Torvald, Linux was intended as a better version of the Unix operating system originally developed at U-Cal Berkley. Linux now runs a large fraction of web servers, and has been gaining some ground in corporate data centers. It has been adopted by some cable TV companies as the operating system running their set top boxes. It has also been used in devices ranging from PDAs to portable phones to TiVo digital video recorders. It has not, however, made any significant in roads in the PC market. At least not yet.

VA Linux was the first company to build hardware systems designed specifically to run Linux. Throughout the 1990s VA Linux was the leading vendor of Linux systems, and enjoyed a find reputation in those data centers where Linux adherents had an impact on buying decisions. Even so, beyond a small number of highly technical people Linux was not widely known, and neither was VA Linux. Then in 1998-99, concurrent with the upswing of commercial interest in the Internet, all this changed. People became enamored of the potential for open source in general and Linux in particular to shake up the software industry. Some people felt that Linux software vendors, such as Red Hat, were well positioned to become the "next Microsoft". This fueled a phenomenal rise in the value of Red Hat shares (the first Linux company to go public) from $20 a share shortly after the August 1999 IPO to an all time high of $150 a share in December of that year (symbol: RHAT).

Just as investor interest in Linux was reaching a crescendo in December of 1999, VA Linux went public (symbol: LNUX). Initially offered at $30 a share, the stock opened at $300 and hit an IPO day high of $320 a share, giving it a peak market value of $11 billion and making it the most successful IPO in history. (To put this in to perspective, around this same time Ford bought Volvo for $7 billion). At the time VA Linux had trailing sales of about $50 million. Investors were particularly impressed with VA Linux because of its excellent pedigree. Linus Torvald was on its board, and many of the top Linux programmers in the world worked for the company. Investors believed they were getting in on the ground floor of what was to become the dominant hardware vendor for the operating system of the future—the Dell of Linux, if you will. They believed that VA Linux, with its big head start, high levels of respect and trust in the hard-core Linux community, top-flight developers and pile of IPO cash were well positioned indeed.

Things unfolded rather differently for VA Linux. For a time the company continued to grow sales at a break neck pace (almost 100% a quarter). Their systems were especially popular among Internet startups. Yet they experienced continual losses. Soon after the IPO most of the established systems vendors announced support for Linux, though with various degrees of enthusiasm. Dell, Compaq and IBM were highly enthusiastic, HP was somewhat enthusiastic, and Sun was grudging. Within six months Compaq and Dell were the leading vendors of Linux systems in terms of number of systems sold. Today, Linux continues to sell well in certain markets, and the biggest vendor in terms of overall sales (hardware and services) is IBM. IBM has dedicated billions of dollars of R&D to developing its Linux business. See this link for more on the role of Linux within IBM today: <http://www.newsforge.com/article.pl?sid=02/01/30/1454220> .

VA Linux has now been renamed "VA Software" following a merger with a Source Forge (a software vendor), and has repositioned itself as software company, having exited the hardware business in 2001, less than two years after it's record setting IPO. As the company put it in their 2001 Annual report: "In business, companies must evolve to address market realities. The market for our systems products was predominantly Internet startups. That market contracted. The overall economy declined. Our sales and margins fell. Our losses rose. To stem those losses, we have exited the hardware business."