Web 2.0 In Depth

Digg This!

I recently read an article in the "mainstream" media that gave me pause. The author made an assertion that the current trend towards Open Source might just be a passing fad. I thought about this and looked critically at the software industry, thinking about whether there was merit in that statement. After all, we have seen plenty of high flyers peter out in a software industry riddled with buzzwords and acronyms-of-the-day. I just don't believe that open source is one of them.

When I worked in the Internet Service Provider (ISP) business, we saw every new consumer Internet fad firsthand. I remember PointCast, a push technology with a screen saver and news ticker that sucked up bandwidth and reduced office LANs to a crawl. I recall the days when every other technical support call was for help setting up the hot new Web browser, Netscape (the code that has, ironically, been given a second life as the basis of the free and open source Firefox), or the time when when thin clients were first championed by Sun Microsystems. However, not everyone grasped the idea that we would have always-on "dumb terminals" and that all our applications would live on the servers carefully maintained and backed up by application service providers (ASPs), though this vision may someday be realized by Google (who may be setting the stage with their Google Desktop (http://desktop.google.com). Thin clients seemed to have found ways to remotely access internal applications in order to keep data from leaving secure networks. Besides, like so many others, I still love my data-laden, fat-client laptop, and as much as I enjoy tools like hosted e-mail and online back-ups, I don't foresee parting ways with my laptop anytime soon.

What makes free and open source software more than a fad? The notion of free software is nothing new - in fact, it's how software was originally distributed. In the earliest days of computing, you usually got the source code for the operating system at the time you purchased the mainframe. At that time, of course, there were very few computers - most of which were for use at large companies, government agencies, and universities - so users often shared programs and source code. Since free software has been around for over 40 years, it can hardly be called a fad. Furthermore, there are more and more long-term projects that continue to gain momentum. Linus Torvalds started Linux, for example, in 1991, and 15 years later, the operating system is enjoying quarter after quarter of double-digit growth in the server market. That's hardly a flash in the pan. Apache, too, the dominant Internet Web server, has been growing by 4.4% since July, and holds 63.09 % of the web server market, according to Internet services company Netcraft (http://news.netcraft.com/archives/web_server_survey.html).

Not only do open source markets have a reliable history, but their future, too, is nearly guaranteed by the continued need for the adoption of open source software in order to sustain the efficiency they have brought to the industry. First, open source software is by and large a pull model. Before any corporate vendor can get involved, software is usually developed to solve specific problems faced by end users who they self-identify in market. Most people like to showcase their accomplishments and share their knowledge (look at the blog phenomenon), and because the development of a piece of software that's useful and well written is something to be proud of, people usually share it. Then, when someone else sees the value of the pet project, he or she is likely to not only use it, but also share improvements that can benefit the original author and the additional contributor. Clearly, it's a more efficient practice than commercial models. Red Hat and Novell, for example, both fund development of the Linux kernel by paying the salaries of developers. They each focus on adding value in other areas as well, like testing and other quality assurance measures, efficient delivery of a commercial version operating system, and technical support. After all, there's little incremental value between the two base operating systems. It is a much more efficient model than say, commercial Unix, where each vendor duplicates efforts in areas that add very little incremental value.

In other industries - especially businesses that have similar needs and whose competitive advantages are based on assets unrelated to information technology infrastructure - shared development can help the whole industry. Unlike commercial software, open source software is driven from the bottom up, which means it's not typically developed by independent software vendors (ISVs) and then distributed. However, we are starting to see that happening, as well as more and more software vendors turning over their software to the open source community, just like database vendor Ingres has done. Some do it because they don't want to maintain the software, but the smart ISVs know that innovation comes from the end user. Mature products can receive new life from the open source vendor community and adapt to an ever-changing IT landscape.

Invariably, leading commercial software in a variety of software segments will be challenged by open source markets. The following are current trends I believe will outlive the fad moniker:

• Virtualization: It's what's hot right now, but there's no slowdown in sight. Over the past ten years, our thirst for faster computers drove the industry to provide blazingly fast chips that sipped more power and threw off a lot more heat. Now, those efforts are being penalized by data centers that can't meet electrical and cooling needs. Virtualization allows consolidation and higher utilization on Linux servers. Commercial software vendor VMware (an EMC company) was one of the fastest-growing software companies in the past five years. The Xen project, a hypervisor, which is a piece of software that allows multiple operating systems to run simultaneously on the same server, is the open source alternative to VMware. Open source virtualization is being advocated by more than just Linux users. Even Microsoft has announced they will be supporting Xen even though it might be competitive with their Microsoft Virtual Server.
• Open source systems management: Systems management is an established software sector used pervasively across many industries that have been paying dearly for expensive seat licenses and up-front royalties. New monitoring companies like Zenoss (www.zenoss.com), GroundWork Open Source (www.groundworkopensource.com), and Hyperic (www.hyperic.com) are offering subscription-based monitoring products that offer very similar products at attractive pricing models. Some of these projects use the same open source projects to build their commercial offerings. Because they are sharing the burden of development, they don't have to fund all the R&D on their own, and they can bring products to market quickly and inexpensively, which enables competition with the Big 4 management vendors (HP, CA, IBM, and BMC).
• Open source business intelligence and planning software: This is more of a prediction than a trend, but it's one of the next sectors to be challenged by upstart open source projects that are replacing traditional enterprise software vendors. It's an industry ripe for open source: long-established vendors who have a broad customer base and a need for customization. Pentaho (www.pentaho.com), an open source business intelligence platform, will start to challenge companies like Business Objects, SAS, and Hyperion. In Enterprise Resource Planning (ERP), the open source company Compiere (www.compiere.com) is going to challenge industry leader, SAP. Both companies will offer a competitive price and an easily integrated solution that will be a viable option to the entrenched vendors.

If you are still skeptical about the longevity of commercial viable open source software, that's fine. There's no place to check the open source scorecard and, truthfully, you could and might conduct all your computing on proprietary closed source software. Of course, you will probably end up with greater expenses from software royalties and a lack of flexibility. As I write this article I am using a mixture of open source and commercial software, but as I look back 10 years ago, I didn't use a single free piece of software as part of my desktop computing regimen. I don't recall even using the term open source or reading about open source in the mainstream media, whereas now, lengthy conversations on the topic are the norm for technophiles and business people alike. Open source has transcended fad status and gone beyond being a trend. It is the evolution of a better way to develop and maintain software, and many of us would be well served to see how we could capitalize on it.