Josef Assad

The advancing stature of open source in the enterprise is unquestionable. Open source technology has a demonstrable foothold in numerous industries and roles, and it has indisputably acquired a portion of mindshare which is appreciating. Looking back to the beginnings of open source where even the terminology was radically different (free software), it can often be difficult to recognize enterprise open source as a direct consequence of the birth of the free software movement. Badgeware as an example was not on the map in the early free software ecosystem, and vital components of healthy open source technology are not on the agenda in many enterprise open source deployment scenarios today.

The consistent and often exclusive focus on cost of acquisition as a competitive advantage is a key factor leading into unsustainable and sub-optimal deployment of open source. It is often the IT budget rather than the longer range IT strategy which provides an entry point for free software in the enterprise. The distinction is crucial: an annual budget at best provides room for tactical initiative, such as cost reduction or intermediate architectural adjustment. The core values defining how technology can serve the enterprise are rarely examined on a 12 month horizon, and it is arguable that the real business advantage to be gained from free software can only be realized through the adoption of the culture surrounding free software technology. For most enterprises, this is a long range proposition.

By advocating free software on the premise of immediate cost reduction, the implementing enterprise never acquires the insight that the adoption of free software introduces new and fundamentally different options for long range IT strategy. This has to be considered a failure since it trivializes the role of information technology in the enterprise: it is not the role of IT to be cheaper, it is its role to be better; and "better" however defined will normally drive at supporting the enterprise's core business more efficiently and more innovatively. This is true across the board: a small importing agency does not have the business objective of improving its ordering system, it has as its objective the reduction of transaction costs. Even software oriented enterprises do not have direct software-related objectives: google does not aim ultimately at having a better search engine, it aims to (for example) place its technology more pervasively and compellingly between you and the information you use. Objectives are never about the technology itself but about the business; selling the cost advantage of open source as a primary strength is a tragic misrepresentation of what the open source option can do for a business.

Efficiency and innovation are not characteristics which are only attainable at a purely technical level. Efficient can mean a faster online order-shipping-billing cycle, and it can also mean faster and more thorough community-driven quality assurance. In the context of an annual technology plan, it may be innovative to introduce mobile banking technology, but in the longer perspective there are forms of innovation which can only be attained through more deliberate and measured embracing of open source as a culture rather than as a simple technological choice. Industries standardize on shared and open platforms in the longer range, and this does not happen without cheerleaders and visionaries. Cheerleaders are not made by building awareness of MySQL as a low-cost alternative to SQL Server.

Enterprises are being sold free software on the premise that these systems can fill certain immediate needs, perhaps that the availability of the source code translates to flexibility, or that the free software alternative is technically superior to current proprietary systems. This is inwards looking, which decisively excludes the enterprise from the greater advantages of claiming membership in community driven technology. The open source model is one which generates a fundamentally different way of creating and maintaining technology, it is not a low cost alternative to Chinese code farms. It is almost an extreme form of outsourcing, where the community determines innovation in a manner independent (if inclusive) of the perceived specific requirements of any one enterprise, it is a technology ecosystem structurally opposed to the emergence of 800 pound gorillas such as Oracle, but friendly to distributed and massively redundant provision of more agile support. The scale of involvement invested in a typical and mature open source project typically outstrips what an enterprise is financially capable of; while this resembles a cost proposition, it isn't. The many eyes approach results in better design just as it results in better code, and a mature product with a solid deployment base will arguably stand a better chance of addressing longer range business requirements (both foreseeable and unforeseeable) simply because it is tried and tested. An SME might build a smaller document categorization system and then face a rewrite five years down the road when there are regional offices and a pressing need for access control; existing open source systems would have handled this changing business requirements environment with grace. Observe how this is a larger requirements proposition, even if it contains multiple aspects of cost advantage.

Reliance on proprietary technology incidental to the core business as a competitive advantage is a highly unsustainable strategy. One distinction which has lost clarity partially as a consequence of proprietary technology business models is the distinction in role of the data from the systems which carry, manage, and present it. The loss of absolute control over the specific technology products inherent in open source migrations needs to be clarified in the context that the organizational advantage is a function of the data, and the enterprise needs to understand the options available to direct the product towards its new requirements. In the past, the IT department needed engineers who could maintain the current system. Under open source, the IT team needs to know how to interact with the community to influence the roadmap, to issue successful code bounties, to know who to sponsor - someone who knows open source and can harmonise community and enterprise direction. It's outsourcing, Jim, but not as we know it.

The sub-optimal marketing of open source to the enterprise also does the technology communities standing behind the specific products a disservice. As the userbase of an open source product expands, we hope for, normatively, a parallel linear expansion of the size of the developer or active user base serving the product. When the decision to deploy an open source system is taken based primarily on cost considerations, it is treated like other non open source systems and the implementing enterprise drops out of the community. There are countless outdated Linux servers out there which serve better as testament to this misappropriation of open source than they do as secure and relevant technology assets to their owners. Superficially deployed open source technology can be actively detrimental and not only sub-optimal.

The principles of open source are not antithetical to the enterprise. The formulation of the expression "open source" was a community response to the understanding that a superficial reading of the classical principles of free software could easily misconstrue it as naive and unprofessional. The rush to open source in business has on the flip side obscured the deeper impact of free software; on a personal level, it changes the way technology permits us to live our lives; on a business level, it changes the way technology is harnessed to create competitive advantage through knowledge. This is not known today to the people building enterprise strategy, and enterprise open source adoption will continue to focus on cost and near term gains as long as organizational long term plans do not speak to how they shall to the benefit of their profit generating activities integrate themselves into open source technology communities, or even to catalyse new ones.