Through 15 years of experience in IT, I've learned the benefits of using open source software (OSS) whenever I can, instead of using closed source products. In particular,
Even though I've had problems with Microsoft, I can't recommend using OSS in every situation. In this installment of a two-part column, I'll not only discuss what I consider to be Microsoft's unethical practices, but I'll offer advice on when not to use OSS. This short series describes my experiences with OSS and proprietary software. In part one, I talked about problems I've had with proprietary software and my take on the total cost of ownership (TCO) debate.
Why it's hard to recommend proprietary vendors
Modern business has a general reputation, not always undeserved, for letting revenue prevail over ethics. Fortunately, most of the people I've encountered in IT take pride in their work and strive to give their clients or employers the best possible advice and service.
As an IT consultant, I help businesses choose software products. More importantly, I advise clients on the best suppliers to use. Commercial software vendors do generally offer good service, but those come at a price, and that price is not just the price tag on the box.
In my IT experience, I've encountered situations in which a commercial vendor, Microsoft, has put its own interests ahead of those of the client. I have experienced the following examples first hand, and this is my short list:
I fail to see how any of these practices can possibly benefit the customer.
So when would I recommend Microsoft to my clients? Practical reasons -- either strategic or technological -- must be considered first. But if they don't decide the issue, then I prefer open source because that development community works on honest principals, starting with making source code available.
Strategic issues push me toward recommending OSS over the products of major commercial vendors other than Microsoft. Businesses need software suppliers that have long-term viability, and commercial Unix vendors and Novell are uncertain bets.
Novell has produced excellent and reliable software since the 1980s, but has also shown a lack of focus, strategy and responsiveness to the market's needs. Currently, Novell is shifting its focus to Linux, but the company's future remains uncertain.
Commercial Unix vendors are generally bound to certain hardware and application software, both of which I've found to be difficult to find and generally expensive. Windows Server and Linux have bitten a sizeable chunk out of the Unix market. As a result, most Unix vendors have been forced to concentrate on niche markets and are struggling.
When not to use open source
Is open source software a cure for all ailments? No. There are many situations in which it is not an option, and there are many possible reasons not to use it in certain situations.
For example, using OpenOffice.org instead of Microsoft Office can be somewhat complicated if you rely heavily on existing macros for Office. Macro compatibility is the one component of OpenOffice.org that still falls rather short of the mark. Be prepared to rewrite most existing macros to get everything to work properly. Another consideration is the type of documents you generally receive from business relations. If these documents rely on macros (e.g., order forms or inventory sheets), you could run into problems.
Open source is also not an option when you rely on back-end applications that have been tailored to work with Microsoft IIS and Internet Explorer. Many business-to-business (B2B) and financial applications suffer from this ailment, and getting them to work with open source software, if it can be done at all, is generally cumbersome and holds no guarantees for the future.
What decision makers can learn from my experiences
In my years in IT, I've seen that closed-source products have practical, economical and ethical disadvantages. Meanwhile, open source products have proven to successfully address these disadvantages in many, although certainly not all, situations.
In real life, I've found that a mix of traditional and open source products is generally the best situation: Use open source where you can, thereby dramatically increasing your ROI and lowering your TCO; and use commercial products where you must, thereby maintaining compatibility with existing applications, B2B systems and business relations.
The deployment of open source products in the corporate environment is not an all-or-nothing decision. You can start gradually by deploying OpenOffice.org on a few desktops here, or Linux and Samba on a file-and-print server there, and seeing how it goes. Such a pilot project quickly demonstrates what works and what doesn't. Be prepared to see great performance in one area, slight changes in functionality in another. Sometimes new approaches will have to be tried to work around differences between products, as you're essentially migrating to a new application.
For now, stick with the well-known and well-supported open source applications. Continuity and expected lifecycle are important, but with the most popular open source products, this will rarely be an issue.
With OSS, your initial investments in time and effort will pay for itself many times over. When properly deployed, open source products will reward you with increased performance and reliability, reduced cost and more flexibility and versatility, thereby improving the performance of your IT department's services as a whole.
About the author: About the author: An IT professional for over 15 years, Frank van Wensveen has worked in systems and network administration and implementation; software development; support; engineering; and consultancy. He currently runs his own IT consultancy and Internet development firm in Johannesburg, South Africa.