Making the Open Source Business Decision

December 28^th, 2002

Recently on Usenet I saw the following question asked about Linux:

At least, Microsoft is in the business of publishing their software. They do it for profit and appear to be profiting handsomely by doing it. How can I be sure that the group of 'volunteer' programmers who wrote this thing won't up and 'get a real job?'

This is certainly a question that vexes some managers who misunderstand the way that Open Source works, and who have a mistaken sense of their business priorities. Faced with this question you need to ask yourself another: "Are you in the habit of making your business decisions based on what is good for Microsoft?" If so, it's a habit you should break. Make decisions based on what is good for your business. Here are some things that I think you'll agree are good for your business:

1. Lowering your operating costs. Obviously, Open Source Software (OSS) does this. Even when purchased, you can install the software on multiple machines. And there is no risk at all of a BSA audit, so you can pretty much do away with expensive software license monitoring.


2. Spreading your risk. It's common sense that you don't "put all your eggs in one basket." Linux is not a one-vendor OS. If your relationship with an OSS vendor turns sour you can go elsewhere. What happens to your business if your relationship with Microsoft goes sour?


3. Surviving the failure of a key vendor. Does it matter ifthose volunteers get a "real job?" Not really. It's happened before and we hope it keeps happening! With proprietary software, the vendor dies and you are left looking for a replacement which is often not suitable. With OSS the project is simply picked up by others.


4. Adopting standards. Linux is built on standards. And what that means for your company is that you're not stuck with proprietary technologies and specialized skillsets. Linux is a UNIX. You can hire anybody conversant with any UNIX and he can sit down and in fairly short order he can be up to speed on your Linux systems. You do not have to go hire some Linux specialist. And the same guy can, with very little training, be as useful on any hardware platform you happen to have. Look at the difference between this and Windows, where Microsoft was ready to de-certify every NT4 administrator on the planet who did not get re-certified on newer versions. How often are you expected to send your people to training? What happens if you don't? Why are the versions THAT different?

Also, keep in mind that volunteerism doesn't mean that the authors of important OSS projects are not compensated. Many OSS projects are subsidized by corporations, organizations, or governments. And some OSS is more professional than you suspect. For example, a key part of IBM's Websphere is Apache. So keep in mind that sometimes when you go out looking for a "professonal" solution, the solution handed to you by the professionals either is or is based on OSS. And often that "real job" that the OSS author got is subsidizing work on Linux. Case in point: Linus Torvalds "got a real job" at Transmeta. He still works on the Linux kernel.

Why have companies like IBM and SGI bet their businesses on Open Source Software? Certainly not because they think it's risky and in danger of evaporating.

The Internet was originally a military project, and one of the design considerations of it was that it be disaster-resistant: with no single point of failure it's very tough to kill. It's designed so that no matter where a bomb drops, the messages are routed around it and they get through. (in reality it can get choked up, but it does recover) Adding Open Source to your business gives you a similar military-grade competitive advantage. There's no single point of failure for an OSS project. Once it's distributed it's got a life of its own. Which means you have negotiating power. You will never be shaken down for payment in advance for software you may or may not use. You will never be surprised by the terms of a contract you couldn't read until after the sale. You will never be shocked by the terms of a contract that were unilaterally changed by the vendor after the sale. You can easily move to a different vendor if you don't like the one you have. You can even become a vendor if you choose. All of this gives you a level of freedom and control over your business that is simply impossible using commercial software.

It's said that businesspeople are afraid to adopt "free software." Not when they know what "Free" is. Businesspeople do not fear Free Speech, Free Will, or a Free Market Economy. They don't need to fear Free Software, either.

Where was Apache before the dotcom bubble burst? Where is it now? Same place. Where was Linux before? Where is it now? Still there! Where were BIND, and Sendmail, and KDE, and the FSF and project after project after project going back to EMACS and the roots of Free Software? Still there. These projects didn't die for lack of money... because they weren't started for love of money. They were started because the industry had a need, and as long as the need doesn't die neither does Free Software.

Now look at commercial efforts. They whooped it up, had a big ol' IPO... money, money, money just falling from the investors' pockets! How many of them survived? Here's the difference: proprietary efforts die even when there is a need. Even when they're making money they can be bought and "integrated." They can be bought by a competitor simply to be killed so as to remove competition. Or they can simply change their minds and move off in a different direction. And it doesn't matter what you paid for when the support's gone. You're the one left holding the bag with no sourcecode, no leverage, and no recourse. Who wants to bet their business on that?