A comment on the adoption of open source by governments and the military:
That process has been underway for almost 16 years. US DoD and allies have been sponsoring internal development via contractors for many years. One of the first SGML hypertext system code bases was purchased by the US Army at that time and is still managed, maintained and developed by USAMICOM and contractors. The reason was simple: transparency of the code base ensures that features are continuously improved and secured against IP threats or malicious code. Transparency is a survival trait. It occurs naturally in certain kinds of systems in certain kinds of ecosystems, but not all.
Cultivation of ecosystems and low-energy transport models are both models of non-linear systems with common evolutionary control types. A cluster of traits acts similarly to attractor basins with an evolution towards different semantic/functional clusters as energy is applied continuously or discretely. The cycles can be changed by local energy applied at strength (eg, a Hohman transfer) or by applying smaller bursts and using force vectors of multiple basins to amplify the change where the force vectors attract, push, or cancel. Maps across such basins model the evolving orbits of any situation-aware system that self-directs. This is a model for trading on time and energy to direct evolution via intelligent choice.
Although the hosting systems (eg, operating systems) could not be secured, reasonable assurance could be provided that this would not be a threat source and in any case, the host systems would not open fast enough to scale, so the persistence of the monopoly systems of the time provided the stable base. The opening effect in these would take a longer cycle but would eventually evolve toward the same attractors that create commodity lagrange points.
Once the switching costs (cost of basin change) begin to orbit closer to markets where commoditization creates lagrange points (minimal switching costs), transparency emerges naturally as the monopolies face the linearly increasing costs of loss leader research. As the logistic curve narrows, so does the available talent. Ideas and costs are inversely proportional. Monopolistic basins begin to lose force in commoditized markets unless they control distribution. The Internet by its very design and therefore, cultivar traits, would enable smaller systems (think messages) to become very difficult to control with respect to impeding distribution. as long as the necessary ideas are available (the flaw of abundance economics, btw).
Like life, the contest is not fair but everyone gets to enter.
If you know your hypermedia history, you know that military applications were the earliest sponsorship source and particularly with regard to markup applications. While there were more public projects such as the Oxford Dictionary work, these were small by comparison. Modern hypermedia emerges for the most part from the deep pockets of the military-industrial complex guided by the deft hand of DARPA, ONR, USAF and US Army research investments.
The Web got the touchdown but the drive down the field was government work.
And that's the problem. Government work is expensive, slow and requires too much planning, management and incentive to be cost effective except in the incubation stages of technologies. Scale achieves simplification and reduction in costs while improving economic conditions. In the beginning of a technology while it holds a defensive/offensive advantage, it is kept close to the emergent economy. When it must scale, it must go commercial.
At the point the Internet was taken out of military logistics stockpiles and made available for commercial exploitation, its function as a command, control and communications was understood to be ineffective in that form. To become effective, it needed to be world wide, integrated, interoperable, and responsive to OSSINT technologies. As the satellite systems had shown, transparency is a key element of peaceful or offensive diplomacy. The technologies for warfare of the late 20th century had become so complex and decisive that only transparency provided a chance of avoiding the inevitable mistake or madness that leads to their application.
There is no perfect solution to technologies evolving faster than the understanding of their application. Technology evolves on a short cycle relative to the evolution of the understanding of strategic deployment and political will. In metaphor, there is a new car every year, but there hasn't been a new model of the user for 6000 years. Like the car, the fastest understanding emerges when the types of application increase (scope is a diverse population) and the reach scales exponentially. That is the basin behavior of commercial systems: it trades on functions for survival. This is a growth strategy; it increases the population but is not necessarily a fitness trait. With function comes complexity and this can weaken sustainability by narrowing choice. Every feature added perturbs the orbit with respect to the distribution.
In commercial systems, the competitive advantage required for survival is holding close control over resources for resources, not the implementations, but the ideas. This leads to IP and to incentives to retain and manage productive talented staffs. This is capitalism. It works as a means to create scaling effects because it evolves competitive traits for creating markets and acquiring market share. The feedback of reward for value is the engine of commercial scale at every level. It has disadvantages because securing competitive advantage reduces transparency required for scaling systems with diverse populations. To offset this disadvantage where it occurs, it is necessary to use the emergent budgets to stimulate cultivar that do exhibit the desirable trait and breed these with suppliers from the wild that have the necessary survival traits.
As hoped, the release of the Internet for public use has achieved that in the emergence of open source systems: a development and business model that mirrors the original goals of Internet communications, survivable communications systems where the locus of control can shift axiomatically if any part of the system is damaged or destroyed emerged to ensure the replication of that system at a rate that ensures an evolutionary stable strategy for sustainability. Open source emerges from an ESS with survivability as its goal. This resonates on the other emergent basin: the user populations.
Transparency across populations invites understanding but it cannot guarantee it. The understanding that diversity is the key to surviving and growing must be as well scaled and reached. Our tools cannot do this for us, but if these systems are as Doug Englebart described them, a means to augment human intellect, they are our tool to apply for growth or destruction. Choose wisely. Fitness is not what is most elite. The elite cultivar does not scale because it does not survive. A surviving system combines the weak traits from the edges of the network continuously. This is why simpler less complex products often evolve by means that are healthier for the ecosystem itself. It is responsive, agile, and adept.
Open source: at this point in history, it is not only logical to do this, it's the plan.
