Following on the heelsoutlining a debate as to whether open source leads to a winner-take-all phenomenon, John Prendergast of Jefferies Broadview pinged me to suggest that perhaps Roy, Dave, and I were missing a critical point.
I asked if I could post his comments, and he graciously acquiesced. He writes:
I've got agree with you here and also with Roy et al. The reality is that both perspectives are valid.
There is a virtuous cycle in open source, where success goes to the successful, making second place an even uglier place to be in open source than in the proprietary world. However, and this goes to your basic point, there are many ways to create value for users and customers, classic differentiation in marketing terms. As a result, we're already seeing a dynamic where later open source market vendors are on a trajectory for success. Interface21 and the Spring Framework come to mind following JBoss as does EnterpriseDB with their implementation of PostgreSQL following MySQL.
What is different in the open source world compared with proprietary models is that open source tends to enforce a discipline of differentiation for later market entrants while proprietary markets have historically been more tolerant of me too copies. The result is that value for users and customers is built out broadly perhaps sooner than would be the case in a proprietary only market.
Isn't that interesting? More copycat companies in proprietary software than in open source. Looking around, it would seem that John is dead on. How many search companies do we have? Database companies (past tense and even now)? Etc. Since each of these companies thinks they're truly doing something different (only because they persist in believing that code is a big differentiator, rather than the fundamental service/value they provide), they tend to spring up in blissful isolation...all 3.2 trillion of them.
Open source software markets are free markets. Efficient. Proprietary software markets? Not so much.