Still on the fence about open source applications? You're not alone. That's why enterprise open source evangelists will be the headliners at next week's LinuxWorld Conference & Expo in San Francisco.
There are good reasons to take a wait-and-see stance about implementing mission-critical open source apps. However, hesitating too long could leave your company waiting in line behind its competitors. Here's what some industry experts have to say about the pros and cons of the waiting game.
The waiting game: Pros
Many corporate IT decision makers, being risk averse and conservative, are slow to adopt new technologies, according to Bernard Golden, said Bernard Golden, president of consultancy Navica Inc. of San Carlos, Calif., and SearchEnterpriseLinux.com resident expert.
"Waiting offers one advantage: not committing to something until it's clear that it will be a dominant trend," Golden said. "You avoid chasing something that turns out to be a dead end."
If chief information officers' business needs are being met, "waiting won't necessarily put them at a disadvantage," said Martin Viktora, CEO and chief technology officer of Kerio Technologies Inc. of Santa Clara, Calif.
CIOs who are shopping for better apps should compare open source to commercial products. "That doesn't mean that they need to rush into hiring new staff and migrate to the latest open source project, especially if their existing infrastructure is performing to their satisfaction," Viktora said.
Seth Hallem, CEO of San Francisco's Coverity Inc., a code-analysis software maker, has seen some shaky open source startups. "Based on our experience working with a number of proprietary software developers and open source projects, we think that if an open source project is new, the 'wait-and-see' approach does have advantages," he said.
Hallem and Coverity have found that time cures what ails a software project. As an example, he points to the quality, security and stability improvement in the Linux kernel between releases 2.4 and 2.6. "Specifically, despite the code size increasing, the bugs per thousand lines of code have actually declined," Hallem said.
Another concern is that new open source products may not have the community backing needed to support enterprise applications.
Some open source apps have communities of thousands of programmers working to improve core functionality and support users, while some newer projects may have only a handful. Implement the product with only a dozen programmers on board, and you could be in for a support "nightmare," said Bjorn Skare, president and CEO of Scali Inc., a Marlborough, Mass., cluster software provider.
Before implementing open source apps, CIOs should take the time to set clear policies on how open source can be used within a company, several experts said.
"We have heard of situations where an employee leveraged open source software for a project, but modified it locally to accomplish a task," Skare said. "When that person left, no one had any understanding of how the software worked. The company thought it was using a standard open source application, when in fact, it now had a unique customer application with all the support issues associated with that."
The waiting game: Cons
Today, there is a relative shortage of experienced open source people. "Waiting means that other organizations will employ the available talent, so late adopters will face talent shortages once they decide to move forward," Golden said. "This will also impact the diffusion of open source skills throughout their organizations, as they won't have open source people available to begin the social dispersion of knowledge."
Organizations that move to open source often free up corporate capital to invest in other IT initiatives. "Companies moving to open source late will see their competition begin to roll out new IT initiatives that late adopters will not be able to afford," Golden said.
Linux provides proof that replacing proprietary software with open source software can cut costs, said Matt Mosman, CEO of Levanta Inc., a management and virtualization software provider in San Francisco. "Linux is changing the economics of the data center," he said. "In some data centers, where proprietary servers and operating systems have been replaced by cheap commodity hardware and Linux, capital cost savings are being reported at up to 50%."
Don't be taken in by claims that open source operational costs offset its capital savings, Mosman said. "The operational cost FUD is based on old news," he said. "The new news is that a lot of new technologies are capable of driving operational costs even lower than on other platforms."
"The disadvantage of not migrating to open source is continuing down that path of reliance on proprietary environments," Mosman said. "This means higher licensing costs, being subject to unruly and frequent updates, and essentially being locked into a restrictive compute environment being dictated by a single vendor."
CIOs who are absolutely sure that their competitors are not using open source apps will be safe playing the waiting game, experts said. Otherwise, "you either lead, or you follow," said Marge Macintyre, program director for IBM Worldwide Linux Strategy.
CIOs who do nothing with open source today could get started just as their competitors are on their fourth or fifth rounds of moving critical enterprise workloads to Linux and/or open source applications, Macintyre said. In her opinion, early adopters will gain strategic advantages, because open source products offer cost savings and business agility.
Summing up, it's obvious that there are risks to adopting open source applications now and to waiting to adopt when they become ubiquitous in the enterprise. To wait or not to wait? That's the question.