This week, for one brief moment in time, my home town has become the center of the open source universe.
No, this isn't egoism talking--it's all about the 6th International Conference on Open Source Systems, hosted at the University of Notre Dame.
This is the first time the OSS conference has been held in North America, having been in Europe after getting its start in Genova, Italy in 2005. Nor is this like any other open source conference you may have attended or heard of: these are academics doing serious research on the properties of open source development, communities, and products.
During the three-day conference, researchers from all around the world have gathered here to discuss their insights on what makes open source tick. To help them, they've invited notable figures in the open source community to speak on various aspects of open source. I've likened it to the penguins speaking directly to those strange folks in parkas just over the ridge, but in reality, it's much more collaborative than that.
The first speaker Monday morning was Larry Augustin, CEO of SugarCRM, who spoke on Commercial Open Source for the Enterprise. Augustin's main point was that the current enterprise software model is broken, given that increasing costs of sales and marketing and reduced costs of product development means customers are paying more for marketing software and less for value.
Naturally, Augustin believes open source is the knight in shining armor: "It has reinvigorated the enterprise software landscape," he told the audience.
Open source, by its nature, cuts sales and marketing costs, because it flips the traditional sales channel model: "now we can let the user come to us."
Development costs are also chopped. Companies or projects still do most of the development work, but now open source development can be leveraged for testing, edge cases, and interfaces.
These are valid observations, and with any other blog, I might call it a wrap right there. But on his way to making this point, Augustin opened up another avenue of exploration that I believe is even more relevant to the datacenter: the five generations of open source software development.
It all began in 1974, when the first generation of open source software (really known at the time as free software) were the hobbyist games, like Adventure and Rogue. They were popular at the time, though many observers thought that was the end of the usefulness of free software.
Along came the second generation in 1982: open source tools, such as gcc, Ld, and emacs. These were certainly very useful, and they did build upon the technology of the games of the previous generation. That, Augustin related, was surely as far as this novelty of free software would go.
Lo and behold, here came the next generation, in the early 1990s: the operating systems. Linux and FreeBSD, among others. More great stuff, but there would be no use for free software higher up the stack, right?
Wrong, it seems. Welcome to generation four: the infrastructure, introduced to the world in 1996. Here was the middleware layer: JBoss, MySQL, and PHP. Open source had surely come a long way.
And, in 2004, the next generation of open source development began with the next higher level in the stack. Applications like SugarCRM, Alfresco, Medsphere, Asterisk, and Compiere, Augustin told the audience, have been the product of this latest generation.
Each of these cycles, Augustin noted, takes about six to eight years, which is just long enough for a new generation of developers to go through college and gain some practical work experience before building really innovative software based on the works of the previous generations.
Given the date on the calendar, we're just about due for the newest generation of open source development. Augustin speculated that it may be cloud computing, as a new generation of developers will likely see the entire Internet as their platform.
This seems a probable, er, development in the history of open source, and if this is indeed the case, we stand on the edge of a new burst of products designed for the cloud. It's an exciting idea, because looking at the innovations in each of the previous generations, there's going to be great software coming soon.