Will Linux and open source computing foster your next enterpriseapplication? Don't take yes for an answer.
One of the great debates about enterprise software revolves around whether Linux and the open source movement will have the same revolutionary potential in manufacturing as this movement has had in the operating system, tools, and server arenas. The latest entry to the debate comes from a company called OpenMFG, whose status as a contender in the "open source takes over manufacturing" debate got a major, and somewhat confused, boost from a BusinessWeek article with the title, "Next from Open Source: Killer Apps?". The question mark at the end of the title begs for an answer, which I'm happy to provide: No.
That doesn't mean open source will never produce apps of any kind. I fully expect open source applications to someday make inroads into highly commoditized functions like general ledger and accounts payable, among others. But will they be killer apps? Probably not. And more importantly, will small and medium-size manufacturers flock to these not-so-killer apps in any great numbers? Not too likely.
The reasons are many; but first, let's clear up the questions regarding OpenMFG. Despite BusinessWeek's misleading headline, OpenMFG is not open source software. Even its developers admit it fails the test. OpenMFG does rely on an open source database (PostGreSQL) and the Linux operating system. That makes it more open source than, say, Oracle E-business Suite (which runs on Linux but uses the Oracle database). But using open source technologies does not an open source application make.
The reason open source won't rival the Oracles and SAPs of the world comes from a mix of cultural, business, and competitive factors. Culturally, the open source community is long on technology but short on deep business acumen. The kind of business process knowledge that is embedded in big enterprise applications has to come from people with knowledge of how the enterprise works. OpenMFG founder Jeffrey Lyon was just that kind of person. In contrast, Linus Torvalds, founder of the open source movement, was a grad student hacker when he created Linux. That legacy lingers on in the open source world.
The business and competitive issues militating against open source killer apps are perhaps even more germane. First is the support issue. The people who use Linux and other open source technology are technologists themselves: this class of software is rightly the purview of the IT department. The people who buy enterprise software are more and more the business users who want great functionality and no-nonsense applications management. They want and need not just competent support, but also a vendor with a direct stake in the customer's success. It's not for fun that enterprise applications customers do a lot of due diligence on their vendor's financial viability. Technology for technology's sake--the imperative of the open source community--isn't what business users want.
The other important reason is that there is a good alternative to open source-like software in the enterprise applications market. It's called outsourcing, and its growing success at Oracle and Salesforce.com will likely blunt the business case for competing open source applications. What Oracle and Salesforce.com are able to do is shift the debate from one of pure feature/functionality to one that takes into account the requirements for implementation, service, maintenance, and upgrade that together cost far more than a mere software license. Getting a free open source license will save on the initial cost. But when it comes to total lifecycle costs, outsourcing looks like a much better deal.
So don't be seduced by the promise of free ERP software from open source proponents. It's not as free as it looks. Open source is making serious inroads in a lot of traditional software arenas, but your enterprise applications will still come from for-profit vendors for some time to come. MA