The Opensource Problem

Hey John -

RE: Opensource

We can see instances where opensource has been used strategically to create unique capability – the perfect example is Google, where all of their platform is opensource. In creating a custom code base, Google developers are four times more efficient than their Microsoft counterparts. Further, as there are no licensing purchases made by Google for proprietary applications, nobody can get a sense of what kind of platform/volume constraints this company might have. They’ve been able to tweak opensource applications to scale the effectiveness of their applications and fine-tune nuanced structures (like file system structures, for example, and I/O performance) based on database optimization, allowing Google to achieve unheard of capability. Yet, they’ve been able to standardize themselves to allow for a reasonable TCO on 200,000 servers in 68 facilities – all using opensource.

Opensource, then, is enterprise-capable in terms of a licensing model and development strategy. Security, as you pointed out, isn’t of issue – it is capability and the ability to execute.
However, we see a problem for the more mundane application of opensource. Google’s strategy is constant evolution – constant improvement upon its platform, which fights Z-Curve returns (more strategic investment in IT yields fewer returns). For Google, they can justify the thousands of man hours in customizing and extending their opensource applications to eek out every bit of efficiency they can. However, where’s the return for a mom and pop running StarOffice (OpenOffice) or a Linux server?

Here’s the value problem for opensource. The initial capital outlay for the licensing may be appealing, but it’s TCO (long term maintenance) is extraordinarly high when compared to licensed product like Microsoft’s small business solutions. For a company to invest in SBS2003, for example, and gained bundled capability like Exchange, IIS, and SQL Server for a $1.5k investment, that is then managed like any other Windows station using component updates and wizards, the _value_ proposition for the small business is much better with licensed software than with OpenSource. Where IT is more of a utility and the Z-Curve needs to be justified, it’s impossible to recommend opensource licensing – the TCO and on-going specialization in maintaining that solution exceeds its value. Therefore, Microsoft’s product offerings look attractive.

What’s nice about opensource is the entreprenural aspect of it – anyone can download, tweak, install, and gain immediate capability, diminishing upfront capital requirements. But there must be ongoing commitment and talent available to extend and maintain these applications. If the business model looks to constrict the growth of IT following startup, opensource isn’t favorable; if the business looks to extend the footprint of IT and rely on IT as a competitive differentiator, opensource makes sense. Naturally, for the common business, you can see a problem here: we’d rather outsource as much as we can when it comes to IT to gain capability but at contained risk and cost, so opensource – again – flies contrary to what’s happening in the industry.

Opensource: good on paper, cool to techies, but poor in execution primarily because of the small business problem. You just cannot convince 80-percent of the economy to shift from purchasing licenses to maintaining software – it’ll never happen. Look for big companies like Google and IBM to extend opensource to competitive differentiate themselves strategically. Look for small startups to leverage opensource to launch into a marketplace. But long term, TCO-contained strategy? Opensource is far from it…

RP Mickler, CISSP MCSE
Mickler & Associates, Inc.
www.micklerandassociates.com

Name (required)

Mail (will not be published) (required)

Website

Notify me of follow-up comments by email.

Notify me of new posts by email.