Open-source software (OSS) was the purview of hackers and other "computer radicals" when I was first introduced to its wonders. Back then, forward-thinking geeks smuggled OSS into corporate America, where it was put to good use for replacing Windows file/print servers and for intranet/Internet infrastructure. Installing OSS had to be a clandestine operation, given that the pointy-hair management types were unwilling to believe that a free product could perform as well as, if not better than, the proprietary software it was destined to replace. Thus, there was little or no management buy-in (pun intended) or approval for its utilization.
Here it is, 15 years later, and the impact open-source software has had on IT infrastructure since then is nothing short of stunning. No longer is OSS held in disdain by stuffy MBAs who couldn't understand the business model and who didn't believe that it was sustainable. Not only is OSS tolerated in corporate America, but it's actually welcomed with open arms. I've had discussions with my peers and clients who have related that they have received queries from upper management as to whether they were employing OSS and, if not, for what reason. So pervasive is the OSS buzz that even the most skeptical observer must have finally come to grips to the reality that is OSS.
The Commercialization of Open Source
Socialism! That's what OSS is. I heard or read that sentiment expressed many times during the mid-nineties, mostly by the MBAs who didn't get it. Perhaps it might appear that way when compared to the business model so long employed by software companies. Even if you agree with that sentiment, you have to admit that the capitalists have embraced the community...and have made a tidy sum doing so. That doesn't sound at all like a socialist system, does it?
While a study of the huge industry that has sprung up to provide support services to companies using various OSS projects could prove fascinating, of greater interest is the effect the popularity of OSS has had on traditional "proprietary" vendors. Some vendors have "gotten religion" and have embraced open source while others are grudgingly accepting that it even exists, at best. It's easy to tell one type of vendor from another. If the company uses euphemisms like "Shared Source Initiative" to try to look like they're embracing the ideas embodied in OSS, then they're definitely in the latter camp. On the other hand, if a company proudly proclaims its involvement in the OSS world in print, word, and deed, then you can be pretty certain that they "get it."
At first glance, IBM seems to be an excellent example of a company that has gotten religion. They certainly trumpet their adoption of the OSS model, at least for some products, but given their long (and sometimes jaded) history in the business, one has to ask the question: Is IBM truly committed to open source?
IBM's formal involvement began in 1999 when it became clear that OSS was more than a flash in the pan. Not only was the media starting to frequently feature Linux (the OSS darling), but reports from the field indicated that instances of OSS were starting to appear at IBM's own customers. Yikes! Like any good competitor, IBM began to examine the threat that OSS posed to its business model. After much analysis, it became clear that, at the time, open-source software development was focused on infrastructure software, such as network services (DHCP, DNS, Web serving, etc.) and, of course, operating systems. IBM's business model was focused more on middleware and application software, so OSS was not perceived as a threat to IBM's main business. Even at that time, IBM was looking to expand into the services market, so a striving open-source market could only enhance its bottom line, not hurt it. When the analysis was complete, it became obvious to the powers-that-be within the company that it would be in IBM's best interests to get involved in the movement, rather than fight it. And get involved it did.
As I mentioned earlier, when IBM got involved, OSS's primary function was infrastructure. IBM saw a great opportunity to get involved with the most overt example, Linux, and get one operating system that could run on all of its hardware platforms. To that end, IBM assigned some of its top developers and significant resources to the Linux environment, thus bringing to bear its great technical expertise in high-performance/high-reliability computing. Linux was already a rock-solid OS by the time IBM got involved, but IBM's contributions took the OS to "enterprise-quality" levels, making its acceptance into corporations assured. Additionally, IBM was able to fulfill its promise to itself of having one OS for all IBM platforms. Linux runs on iSeries/System i, xSeries, pSeries, and zSeries, so that puts IBM into an excellent position. Any of its software products that run on Linux can theoretically be ported to any of its hardware platforms, making them much easier to sell across the complete spectrum of customers it currently has or acquires.
When I say significant resources, I mean significant. Many of you might remember that on December 11, 2000, then-IBM CEO Louis Gerstner announced at the eBusiness Conference and Expo in New York City that IBM would spend $1 billion on Linux in 2001. Others within the company explained that IBM had already spent close to that amount from mid-1999 up until Gerstner's speech. Two billion dollars is not chump change, even for a company as large as IBM. They didn't wade into OSS; they dove in.
Besides Linux, IBM joined the Apache community and started donating to the Apache HTTP server project. It embraced that software, making it an integral part of its own WebSphere product. I'm sure that many readers remember when IBM ported Apache to the iSeries, replacing the traditional HTTP server (I'll call it "Coke") with the Apache version ("New Coke"). Unlike the soft-drink company's ill-fated venture into tweaking an existing product, this one succeeded spectacularly, therefore averting the need for a "Coke Classic." IBM's old HTTP server has been supplanted by the OSS version, cutting IBM's development costs while providing to the customer what I consider to be a superior product.
IBM has not limited itself to participating in existing OSS projects. It has steadily been open-sourcing parts of its own technology portfolio, especially in areas not likely to affect their bottom line, such as development tools. Almost a decade ago, IBM released the source code for Jikes, its Java compiler. It also contributed some of its "Cloudscape" technology to the community, which became the basis for the Java database called Derby (a project within Apache). While you may have only passing familiarity with these two projects, you probably know about Eclipse, the feature-rich Java IDE, even if you don't do programming on a regular basis. Even as recently as last month, IBM has been open-sourcing its own products as it announced that it is opening up its collaborative development platform, known as Jazz, to the open-source community. (Jazz is part of Rational software, the integrated tool set for improving the software development process).
To get a real appreciation for the depth of IBM's involvement with OSS, you really need to visit IBM's OSS Web site and spend time perusing the various projects they have their hands in. I haven't even scratched the surface with the few that I chose to list because of their visibility.
The Nuclear Option
Any discussion of IBM's commitment to OSS wouldn't be complete (nor could an absolute affirmative be given to the title question) without discussing the dark side of the open source world: patents. While OSS was relatively small and flying under the software giants' radars, no one seemed to care about the pedigrees of the various software components. Once OSS climbed into radar coverage, the old guard started to take an interest: This stuff could start eating their lunch! The affected companies had to decide how to handle the threat. We've already looked at the path Big Blue chose to follow, and you can be very sure that they chose this path only after a thorough examination of the business and legal ramifications.
The tack that the "other" software giant chose (where lock-in is virtually a requirement if their business model is to continue to be successful) was to attack OSS using the weapon that they've employed with such great success for so long: FUD. Soon after Linux (and OSS in general) became threatening, the skies started to fill with FUD missiles. The early ones had tiny warheads carrying innuendo regarding the reliability, stability, and security of OSS. Later FUD missiles had greater firepower and appeared to be launched by mercenary armies (I refer to the SCO v. IBM case). One of the biggest FUD warheads delivered the message that free software infringed on over 200 patents held by the attacker, giving pause to the overly cautious IT manager for deploying OSS (which was the desired effect). It's been almost a year since that bomb exploded, and we're still waiting for the specific list of patents that are alleged to be infringed. I don't think that we're ever going to see any attempt at enforcing these alleged patents; that would be tantamount to using the nuclear option and would ultimately result in a round of "show me yours and I'll show you mine," which could be embarrassing.
As a demonstration of their commitment to OSS (and to ameliorate the affects of the patent FUD missile) IBM recently pledged non-assertion of patents against open-source software for 500 of the patents that they hold. This ensures that neither future OSS developers nor consumers of the products will need to worry about the technology that the patents cover, while at the same time throws the gauntlet down to other companies to follow suit. That move is a definite win for the OSS community.
Yes, They Are Committed
So, is IBM really committed to open-source software? I think that we can safely say yes. The company has already invested huge amounts of research and development money developing OSS, and it continues to do so. The company has also invested huge amounts of money fighting the SCO lawsuit, having set legal precedent to establish the pedigree of Linux. (SCO is virtually nothing but a bad memory now). And IBM has bet the farm that OSS will survive by developing its flagship products to utilize or run on it.
Has the OSS strategy paid off? If you review IBM's earnings for the fourth quarter of 2007, you'll note such statistics as "Total Global Services revenues grew 17 percent (10 percent, adjusting for currency)," "Revenues from the Software segment were $6.3 billion, an increase of 12 percent (6 percent, adjusting for currency)," "Revenues from IBM's middleware products...were $5.0 billion, up 13 percent versus the fourth quarter of 2006," and "Revenues from Rational Software...increased 22 percent compared with the year-ago quarter." All of these business entities are in some way affected by IBM's own OSS efforts, and all seem to be thriving in spite of the current downturn in our economy. Do you think that the reports would be as rosy if had IBM stayed out of the OSS community? I don't either.