I spent October 25 and 26 in Toronto attending the FSOSS 2007 conference put on by Seneca College. I didn’t attend the conference primarily to hear the conference presentations; my main aim was to talk with the people associated with two projects that the Mozilla Foundation has funded, namely the Mozilla-related educational activities at Seneca College and the Mozilla-related accessibility work at the University of Toronto’s Adaptive Technology Resource Centre.
However as it turns out some of the presentations I saw had interesting connections with my Seneca and ATRC discussions. In this post I’ll give my thoughts on how Seneca’s efforts relate to the broader world of business and education. (Note that these are my personal opinions only, and do not necessarily reflect the positions of the Mozilla Foundation or Seneca College.)
Open source businesses and the value of commit access
The closing keynote presentation at FSOSS 2007 was by Dirk Riehle of SAP Labs. If you don’t have time to watch the video you can get the essential points of Riehle’s talk from his original paper and a recent interview he did for Datamation. Riehle’s thesis has several parts, including some claims about software industry economics, closed source vs. open source business models, and so on. However in relation to this post Riehle’s two most important claims are as follows:
The move to open source negatively impacts the job prospects of the average software developer, but improves the prospects of developers who are core contributors (“committers”) to open source projects.
Job and career prospects for committers participating in “community” open source projects (“software that a community develops”) are better than those for committers participating in “commercial” open source projects (“software that a for-profit entity owns and develops”).
Riehle’s contention is that the lowered barriers to learning about and participating in open source projects mean that there are more developers with experience with such software. Thus the typical developer within an open source project will encounter relatively intense competition for job opportunities relating to that software. But if a developer advances within a project to become a core contributor and gains commit access then they will be more attractive candidates, particularly to companies with an interest in having influence within that project.
According to Riehle this dynamic works best in community open source projects (e.g., Linux, Apache, Eclipse) where there are multiple companies that can gain advantages from employing committers, and hence competition among companies to recruit the relatively scarce candidates. In commercial open source projects (as defined by Riehle) a single company typically employs all (or nearly all) committers, so bidding wars to hire committers are minimized.
Creating Mozilla contributors
What does all this have to do with Seneca? Many traditional universities and colleges are research institutions; two of their most important products are published papers and the people who produce them. On the other hand Seneca is (to quote from its full name) a “College of Applied Arts and Technology”; its goal is first and foremost to produce graduates who have marketable skills. Seneca’s Mozilla-related activities clearly support this goal:
Based on various reports from existing Mozilla contributors and others, the students in Seneca’s Mozilla- and open source-related course appear to be getting valuable hands-on experience in what it takes to participate in the Mozilla project, whether that be as testers, build engineers, developers, or whatever. In effect Seneca is training a cadre of people who likely have a much better chance of becoming core Mozilla contributors than typical computer science graduates. If we accept Dirk Riehle’s thesis (and it seems plausible) then these Seneca students will as a consequence have correspondingly better prospects for lucrative employment, whether at the Mozilla Corporation or one of the other companies looking for Mozilla- and Firefox-related expertise.
It’s worth noting here that although the Mozilla started out as a corporation-initiated and -sponsored open source project and is still dominated by a single organization, in practice it has many aspects of Riehle’s community open source projects. (Why this should be so is a topic I’ll skip in the interests of brevity; however I think the non-profit nature of the Mozilla Foundation is a major factor.) In particular, although the Mozilla Corporation is still the largest employer of Mozilla developers there are now several other companies that have a vested interest in having some influence over Mozilla development, and that Mozilla developers and other contributors might consider as potential employers.
Seneca College and disruptive innovation in higher education
At FSOSS 2007 I had a conversation with David Humphrey and Chris Tyler of Seneca about additional things Seneca College could do in the Mozilla and open source spaces, and how the Mozilla Foundation might help Seneca achieve such new goals. Since our ideas about that are still in flux, I’ll skip any detailed remarks on that topic for now. Instead I’ll discuss a different topic, namely, why Seneca College? Sure, Seneca is in Toronto, and so are a lot of Mozilla folks, but to be honest although I was previously aware of several Canadian academic institutions I’d never heard of Seneca prior to its Mozilla activities. If it’s a simple matter of proximity to Mozilla people then there are a lot more Mozilla people in Mountain View than in Toronto, not to mention world-class institutions like Stanford University and UC Berkeley.
So what’s going on? I think this is a perfect example of “disruptive innovation” acting in the education market, and as it turns out Clayton Christensen provided a helpful guide to it in his book Seeing What’s Next. In this view institutions like Stanford and Berkeley are focused on “sustaining innovation”, i.e., improving the “product features” for which they’ve been traditionally recognized and valued by their most demanding customers: a comprehensive computer science curriculum, leading-edge research programs, and tight relationships with local entrepreneurs and venture capitalists that can be leveraged for the benefit of the institution and its students. (This last advantage can be had even without completing one’s education, hence the “college dropout turned Silicon Valley entrepreneur” phenomenon.)
Elite institutions like Stanford and Berkeley are doing fine by their own lights, and we’d therefore expect truly disruptive innovation in higher education to occur not at those places but rather at institutions on the margins. Christensen notes three trends in education that represent disruptive innovation at work, and all three are relevant or potentially relevant to the case of Seneca:
The first trend is the growing popularity of community colleges and other institutions that offer cheaper alternatives to traditional four-year degree programs. This trend, of which Seneca is a part, is an example of “low cost” disruption: As traditional degree programs become increasingly more expensive, students look for other options that are less expensive but still good enough to serve their own personal goals, which are often more limited than those of students who attend elite institutions. (Not everyone aspires to be — or is cut out to be — a famous and wealthy entrepreneur; many people would just like a well-paying job doing work that’s personally satisfying.)
The second trend noted by Christensen is the growth of for-profit institutions (like the University of Phoenix) that are targeted at older part-time students and make heavy use of distance learning using the Internet. This is an example of “new market” disruption, bringing additional higher education to people who traditionally did not pursue it. (Christensen’s phrase for this is “competing against nonconsumption”.) Seneca College is a public non-profit institution (chartered by the province of Ontario), and to the best of my knowledge Seneca’s Mozilla-related activities conform to the more traditional model of young students receiving on-site instruction. However it’s perfectly possible to imagine people already working in the IT field who might want to get an entree into new and exciting areas like open source development, and due to its distributed and decentralized nature open source development should be a good match for Internet-based learning.
The third (and I think most interesting) trend noted by Christensen is the rise of “corporate universities” providing specialized training to a corporation’s employees, including such prominent examples as Motorola University, GM University, and GE’s John F. Welch Leadership Center. Christensen discusses this trend (another new market disruption) mainly in the context of management education, with corporate training serving as an alternative to traditional MBA programs:
Traditional business programs excel at training managers in general business theory and exposing them to a diverse network of business leaders. But they are ill-equipped to provide learning customized for an individual company or individual employees. Corporate training programs lack campuses or access to a high-powered alumni network outside of the company. But modular, customizable corporate training has an advantage that independent M.B.A. programs can't match — a product specifically designed for each employee's needs (Seeing What's Next, p.112)
By analogy, Seneca is offering an alternative to traditional general-purpose computer science programs based on the ACM/IEEE curriculum, an alternative that is tailored to the particular task of teaching students to be productive contributors to an open source project (in this case Mozilla), thereby increasing their value to employers participating in that project. Seneca is thus in effect serving as an outsourcing provider for a prototype “Mozilla University”.
Note that this “open source university” idea is even more powerful than the typical corporate university discussed by Christensen, at least as it might be applied to Riehle’s community open source projects. For example, graduates of GM University no doubt learn many things that would be useful were they to leave GM and go to work for another auto manufacturer. However they’d still likely face a learning curve getting up to speed on the “Toyota way” or the “Ford way” as opposed to the “GM way”; this encompasses both learning explicit knowledge about corporate products and processes as well as acquiring more tacit knowledge that can be gained only through active participation in the corporate culture.
On the other hand, since Riehle’s community open source projects (Linux, Apache, Eclipse, etc.) are cross-company by definition, a graduate of (say) a hypothetical “Linux University” would likely face minimal barriers moving between any of the companies involved in Linux kernel development; ditto for Apache, Eclipse, and (as noted above) Mozilla itself. This is especially true if the curriculum required students to invest a substantial amount of time in actively participating in the project in question (as Seneca’s does).
Asymmetric competition in higher education
As noted above, Seneca College and similar institutions are still bit players in a world of traditional computer science programs dominated by the elite research institutions. But need this always be the case? One of Clayton Christensen’s key concepts is “asymmetric competition”, in which new entrants to a market take advantage of opportunities relatively ignored by incumbents and in the course of doing so can grow to eventually threaten the dominance of incumbents. (Christensen uses the metaphor of new entrants being defended by the “shield of asymmetric motivation” and learning to wield the “sword of asymmetric skills”.)
Programs like Seneca’s demonstrate possible new ways to teach software development: in the context of large real-life projects, with an emphasis on the social and community aspects of projects, and through an inter-disciplinary approach that teaches not just programming but also other skills such as QA and testing, release engineering, user-centered design, evangelism and marketing, and even software as a business. As open source development and related trends (e.g., agile development) take hold, Seneca and other institutions that might emulate it could find themselves more aligned with the needs of both students and industry, and more successful as a result.
As noted above, the elite institutions have no real reason to be threatened, at least not in the next few years; if nothing else their importance as hubs of business networks will ensure their relevance. As they move up market Seneca and others like it are much more likely to threaten the dozens of conventional computer science programs at second-tier institutions. (One sign of this is Seneca’s move to offer a four-year “bachelor of software development” degree.)
However in the long run Seneca or another similar institution could potentially break into the ranks of elite institutions, albeit an elite that might be based on somewhat different criteria than today. After all, if an institution like Stanford, originally founded as a “avowedly practical” institution (in contrast to traditional east coast universities), could ride the wave of post-WWII technology and business advances to become the “Harvard of the West”, who’s to say that Seneca (or a Seneca imitator) couldn’t aspire to some day be the “Stanford of open source”?