Open Source By Pedro Jaime

Open Source


From an investment research perspective, the term "open source software" refers primarily to a set of terms and conditions associated with the software's license. There are dozens of such licenses but those designated Open Source Initiative (OSI)-compliant meet a set of 10 characteristics outlined on the group's web site that capsulize the open source philosophy.

The term "open source" derives from the fact that the software's source code is at least available and in fact ususally distributed with the executable code (the normal way of distributing software). Per one of the typical terms and conditions in an OSI-compliant license, the software can be (1.) "freely" changed and (2.) somewhat freely redistributed (depending on which open source license is used).

Open source software is not the same as free software and the term free software does not necessarily mean available "at no charge." There is free software that is not open source and vice versa. See the Free Software Foundation (FSF) web site ( for more details. This issue is less relevant to investment research because very few public companies embrace the FSF philosophy.
[edit] Open Source Overview

Open source software development is relevant to investment research because it reduces the research and development expense of the publicly traded companies that make use of it. Much use of open source involves burying a component or software artefact deep into an otherwise closed-source software product or service (being careful that the open source license permits that use). For example, some version of the Apache HTTP web server is a component of most leading closed-source application servers. Alternatively many publicly traded software publishers and service providers redistribute a discrete piece of open source software instead of developing their own software for that function. Both approaches save the public software or services company from reverse engineering or otherwise creating from scratch a particular commodity function, or licensing the software from a closed-source software publisher.

Otherwise the development aspect of open source is not especially relevant to investment research. For development, open source software in theory relies on a community of independent contributors, some of whom are volunteers and some of whom are employed by technology companies. In reality, most popular open source software is written by employees of the commercial entity most assoicated with it (e.g., Fedora Linux with Red Hat, Geronimo application server with IBM, LogicBlaze message-oriented middleware with Iona/Progress). There are exceptions (e.g., Drupal content management software and Acquia) but the tendency is to merge the project with the company as the company matures.

There are a variety of open source software foundations that are more independent than the norm from the companies that use their project code for revenue generation. However even these foundations are heavily funded by technology companies (e.g., Apache Software Foundation by all the major market participants including Microsoft, Mozilla by Google, the Linux Foundation by IBM and HP, and so forth)

Any individual can contribute any code in any area they prefer, though there is generally a pyramid management structure that ensures that aspects of the software are developed in a rational fashion and determines which code/modules are eventually included in future versions of the software.

New versions are typically released much more frequently than closed-source software. With a large body of developers and a frequently updated product, open source software has been championed as a way to build better quality software faster and at a lower cost. This also is theory.
[edit] Open Source Platforms

The most well-known open source software is Linux, a UNIX operating system derivative named after its creator, Linus Torvalds. The Linux Foundation, where Linus works, is more or less "in charge of" the operating system kernel. (Of course under open source development philosophy, anyone can take it and fork it into something else but Linus owns the name Linux.) The Linux kernel is turned into a platform by adding a wide range of utility software, primarily developed under the banner of the Gnu organization (, which is an adjunct to

The combination of the Linux kernel and the Gnu utilities is still very primitive as compared to the way typical enterprise operating software such as IBM AIX or OS/400, HP VS, ormore recentlyMicrosoft Windows Server has been distributed for use for 50 years. So one or more open source projects were formed to combine the various components into a more enterprise-friendly distribution. The most well known of these efforts is Debian. Commercial entities were also formed to productize the combination, These include Red Hat, Suse (later acquired by Novell) and others. In kind of a hybrid of this process of moving up the technology stack, Canonical productizes Debian.

The most widely used open source software however is probably the Apache Software Foundation's Apache HTTP web server. Linux/Gnu and Apache HTTP use very different open source software licenses. (It is the Apache license that allows it to be bundled into closed source software in such a way that the resulting product does not have to be open sourced.) Ironically, more Apache HTTP software is probably running on Windows than Gnu/Linux operating platforms. (In fact, according to a census project begun by open-source-software service provider OpenLogic in 2008, most open source software runs on Windows. Note however that the census is not necessarily statisitcally accurate because it uses an opt-in survey methodology.) The Apache Software Foundation also sponsors dozens of other middleware and application projects, almost all of which are associated with some commercial entity.

Although there are tens of thousands of open source projects underway apart from Linux and Apache, there are only a few that have gained widespread adoption at the enterprise level (and therefore are of particular interest to investors). The others include the MySQL databases (acquired by Sun), the Mozilla Internet browser, the JBoss application server (acquired by Red Hat), the Xen hypervisor (acquired by Citriz), and a series of programming development tools such as Ruby, PHP, Python and Perl.

Almost none of the above is directly relevant to investment decisions unless a publicly traded company's primary revenue stream involves the open source project's code (and therefore must adhere to the OSI-compliant license characteristics). The only such company currently is Red Hat.

Of most interest to investors, many software products originally developed and marketed with a more traditional perpetual or periodic right to use license for a fee (and with closed source code) such as Java Enterprise Edition (JEE), Sun's Solaris operating system and StarOffice collaboration software (open sourced as OpenOffice), the Compiere ERP suite, and the Ingres database have been made open source retroactively, typically after finding less than desired success when marketed traditionally.
[edit] How do companies profit from open source?

A number of companies have been built around a so-called open source business model, where they license their software for free or very inexpensively, but derive their revenues from support and training services. This model is really no different than most other long-time software companies which almost always sell subscription maintenance service after selling the initial right to use license. In fact, most long-time participants in the software market realize well more than half of their software-related revenue from this source.

Many companies following the so-called open source business model adopt a tiered approach to their product offerings, releasing a completely free community edition of their software, and then charging a license fee for their enterprise edition, which is often based on their community edition but includes additional features or is more highly scalable. Again, this marketing tactic is not different from many publishers of closed-source software.

The winners so far in the open source game would theoretically be those companies associated with the most popular open source projects but only one—Red Hat (RHT) is an independent public company. While SourceForge (LNX), a popular hosting site, publisher of open source related web sites, and code repository for open source software, is a publicly traded company, its revenues are not derived from the sale of any open source products. Most other companies associated with popular open source projects (e.g., Gluecode, LogicBlaze, MySQL, Suse, Xensource, Zimbra) have been acquired by large long-time software suppliers.

From an investment perspective, it is these large long-time software suppliers that are the real winners because they save on R&D expense as described above and are able to more quickly get to market or buy market share. Sun (JAVA) is a special case because it retroactively declared much of its software open source in 2006 but there is little indication as of September 2008 that that decision has improved Sun's fortunes as a public company.

Many enterprise users of IT have been winners because the open sourcing of commodity software such as application servers, web server software, development tools, and other middleware has driven down the price of middleware in general. It has also forced traditional middleware suppliers to add more features/functions in order to to differentiate themselves from the open source commodity software. (In reality, from an investment prespective there are very few independent middleweare companies left anyways.)

This same commoditization phenomenon is likely to overtake the market for enterprise applications but that market traditionally lags the middleware/tools market by as much as a decade so it is still unclear if open source will have much effect. The enterprise applications market is much more likely to be roiled by the Software as a Service (SaaS) trend, which has a peripheral association with open source.
[edit] Open Source Winners

Red Hat (RHT) and Novell (NOVL) are the two largest Linux vendors. These companies each market their own versions of Linux, which they test for reliability and ensure interoperability with a variety of popular applications. Each has a community associated with it to actually update their Linux version; Red Hat's group/project is called Fedora and Novell's is called OpenSUSE. With analysts predicting that Linux OS servers will continue to gain market share from UNIX in the future, these companies are poised for success.

Oracle (ORCL) also markets a Linux product, based on Fedora, and claims the effort is successful. Red Hat claimed in September 2008 that it sees very little evidence of Oracle in the market but that is typical of competitors underestimating Oracle. Oracle users are so loyal that they do not often even call in other suppliers when they are acquiring new technology.

Another popular Linux distributor is Canonical whose group/project is called Ubuntu. It is not a public company but its founder Mark Shuttleworth has previously taken companies public and made the fortune that allowed him to fund Canonical (and ride a Soviet rocket into space) selling an Internet-related company to Verisign in late 1999, right before the dot-com bubble broke. He shows every sign of being into Canonical for the long haul.

It should also be noticed that Novell has a major partnership with rival Microsoft that includes a reseller agreement by which Microsoft distributes subscriptions to Novell's SUSE Linux software, and the two have joint research and development operations. Under the agreement Microsoft agrees not to file patent-infringement charges against Suse Linux users if it turns out Linux violates any Microsoft patents. Under this agreement, ironically Microsoft may soon become one of the largest distributors of Linux.

As noted above, from an an investment perspective, it is the large long-time software suppliers that are the real winners because they save on R&D expense as described above and are able to more quickly get to market or buy market share. These include IBM, HP, SAP, Misys, and many other companies.Sun (JAVA) is a special case because it retroactively declared much of its software open source in 2006 but there is little indication as of September 2008 that that decision has improved Sun's fortunes as a public company.

Yahoo! (YHOO) and Google (GOOG) are reportedly large users of open source software. Popular websites like these generally own or collocate large server farms that handle all of their web traffic. The ability to maintain hundreds of servers using inexpensive open source software is a huge cost saver for these companies because they don't have to pay per-server license fees which are often the hallmark of proprietary solution pricing structures. Yahoo also owns Zimbra.

Equinix (EQIX) and SAVVIS (SVVS) are two large providers of collocation, hosting, and data center services. These services require the use of hundreds of servers, and thus these companies similarly benefit from inexpensive open source server technology.

Akamai Technologies (AKAM) employs over 15,000 servers to help deliver website content faster throughout the world. They too save money in maintaining open source servers.
[edit] All Traditional Software Suppliers Jump on Bandwagon

Although many closed-source software providers initially hesitated to embrace the open source development methodology even though it could be demonstrated that it reduced their R&D expenditures, that is no longer true as of September 2008. It can be argued that open source has taken over the traditional software industry or the traditional software industry has taken over open source. Either way, open source development/license issues should have very little impact on investment decisions.
[edit] No one is hurt by the growth of open source

It is hard to specifically identity any "losers" because license terms and conditions (which is really what open source is all about from an investment perpsective) have minimal effect on market dynamics.

Microsoft is often considered a loser because all the companies that migrated outdated Unix servers to Linux could have been migrated to Windows Server. But because Linux is a UNIX derivative, such a migration would have been unlikely anyways. In addition, as noted above more open source software runs on Windows than any other operating system and this is possibly business Microsoft would not have received without the open source movement. Microsoft has steadly gained both server-software and overall-software market share and more than quadrupled in revenue since Linux was first aggressively supported by companies such as IBM. So it is hard to consider Microsoft a loser.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License