Open Source TCO: Total Cost of Ownership and the Fermat’s Theorem

Gartner’s 2009 predictions have been widely commented over these days, leaving space and opportunities to rediscuss the open source mantra “we cost less”.

Migrate to open source seems the cheapest solution, at least on an individual basis, but enterprise migrations are not an easy game to play, and TCO doesn’t look like the ultimate answer.

Any similarity between the Fermat’s theorem and the cost benefits of open source?

Simon Phipps apparently denies that the cost benefit is the main attraction, Savio Rodrigues echoes mentioning a third party TCO study comparing open vs proprietary solutions. The report found that the cost of administering and managing the open source product outweighed the upfront license cost savings.

Are they wrong?

Look at the numbers!” by David Wheeler or some Gartner recommendations go, at some extent, in the opposite direction, but again TCO calculations are prone to be defective and controversial even.

Open Source Self-Sufficiency as well as the ability (or inability) of global solution providers to offer multi-year contracts to “migrate, innovate and operate” have to be carefully considered.

Roger Burkhardt predicts:

Systems integrators have the financial strength and pricing methodologies to propose fixed price multi-year contracts for migration and operations to their customers and in many cases to fund innovation as well. These offerings will cover a spectrum — from migration to a fully managed service, to migrations followed by on-premise support and maintenance services.

Moving from artisanship to industrial can be done, must be done. A broad spectrum of open source opportunities will be lost otherwise. Open Source companies have to bring solution providers and big system integrators in the equation, open source ecosystems are mandatory.

People, Processes and Products correlate to productivity in sophisticated systems, keeping it in mind we might end proving that open source products have a better ROI, though.

It took more than three centuries to proof that the equation anbncn has no solutions in non-zero integers a, b, and c if n is greater than 2.

Will it take less proving that open source is cheaper?