Moving Beyond the Linux Beachhead
Five open source trends that are impacting business
By: Bernard Golden
Jul. 27, 2004 12:00 AM
What began as a small movement based in a Finnish student's apartment has mushroomed far beyond his or anyone else's expectations. Linux, sprung from Linus Torvalds' imagination, has alighted in countless data centers and now stands tall as the software strategy centerpiece of technology giants like Hewlett-Packard and IBM.
Despite bad-mouthing and the sowing of fear, uncertainty, and doubt, Linux has successfully established a beachhead in IT organizations. Cost savings and control have made it an acceptable choice in a variety of organizations, from Google to Sabre Holdings (corporate parent of travel site Travelocity). However, open source is about to move beyond this beachhead and drive further into IT territory. As the saying goes, "You ain't seen nothin' yet."
Over the past several months there have been a number of open source developments that carry critical implications for organizations that rely upon software to run their business - in other words, every organization. Open source is moving into new positions in the software stack, being distributed by both new and venerable companies, and being adopted and modified by new industry initiatives. Every IT organization will need to be aware of and respond to these open source developments - and to do so, new IT skills will be necessary. Following are five open source trends that will dramatically impact your business.
Open Source Continues to Grow Dramatically
While many people believe that open source begins and ends with Linux, there are actually more than 80,000 open source projects available for download at SourceForge (www.sourceforge.net). This number is up from about 70,000 less than a year ago. This statistic illustrates two things: the large numbers of open source products available for use, and the increasing numbers of open source developers.
The enormous variety of products available under open source license demonstrates one of the key issues of open source: the challenge of managing such an embarrassment of riches. As IT organizations consider adding open source products to their software stacks, they will have to decide which products should be used and which should be avoided. With so many to choose from, it will take more than a Ouija board to help make the right decisions.
The number of new projects also indicates that increasing numbers of software engineers are becoming involved in open source development. One of the typical concerns IT organizations have about open source is whether a given product can rely on volunteer labor. The growth in projects indicates that open source continues to attract new participants.
Open Source Endorsed Further Up the Software Stack
Many IT organizations have been reluctant to embrace open source without endorsement from well-respected technology partners. The rise of Linux has been immeasurably aided by the endorsements of IBM, Novell, and Hewlett-Packard. With their vouchsafing, IT organizations have aggressively moved forward with installing Linux boxes throughout their infrastructure.
Several of these partners have now put their stamp of approval on other open source products that live further up the software stack. HP recently announced its support for the open source relational database MySQL and for the open source J2EE engine JBoss. CA recently endorsed JBoss as well. (Please see Figure 1 for a depiction of a prototypical software stack and where open source products can fit in the stack.) These endorsements mean that IT organizations can rely on trusted technology suppliers for further open source initiatives.
However, the broad coalition of Linux supporters does not necessarily apply to these new products. As yet, Novell has made no formal announcements regarding MySQL or JBoss. IBM is less likely to offer as strong an endorsement of them, given its own offerings in these product areas. This means that IT organizations interested in exploring new products for their software stack will face additional questions about whether their technology partner endorses those products.
Open Source as the Foundation of a Business Model
One of the most difficult tasks any new software startup faces is distribution: gaining market awareness and getting products used by real customers. Conversely, open source is easily distributed - available for download at the click of a mouse, with no "Approved Vendor List" or tedious contract negotiations to stand in the way of installation and use.
Some inventive entrepreneurs have looked at this distribution mismatch and decided to adopt open source as the foundation of their business strategy. Bypassing the entire long and expensive sales cycle, they rely on open source distribution to build a customer base that can provide other sources of revenues. Following are several examples.
GroundWork provides "turn-key, open source-based IT infrastructure management solutions." Its products enable IT organizations to track the availability and performance of their networks, systems, and applications. GroundWork provides the monitoring system at no cost, with additional management and analytical modules available under commercial license. This can be thought of as a razor-and-razorblade strategy: GroundWork gives away a software product and sells one or more other products that make the base product more useful.
SugarCRM is a new company offering open source Customer Relationship Management software. The founders come from commercial CRM vendors and have concluded that they can build a viable services business based on free distribution of CRM software. Since CRM software implementations typically require significant customization and integration, bypassing the license opportunities in order to focus on service revenues offers great promise for both SugarCRM and its customers.
Compiere is an established open source ERP/CRM product. With over 700,000 downloads, Compiere is now being used as a foundation project for other open source projects that extend its base functionality. The product is used in many countries, with localizations available in a large number of languages. Compiere is developed by a commercial entity, ComPiere, which realizes revenues through training and support.
One interesting aspect of these companies is that they focus on an area that has been a complaint about open source in general: applications. Most open source projects deliver technology infrastructure software, but very few have offered business application functionality. These new entrants offer end-user functionality and position open source even further up the software stack. It remains to be seen how viable these companies are as businesses, but unlike most startups, their customers are not nearly as vulnerable to supplier issues given the open source nature of their products.
Products Revitalized with Open Source Strategy
Two well-established technology vendors have recently made the news with announcements that they plan to offer existing products under open source licenses. Sun has announced that they plan to offer Solaris as an open source product. Sun has been suffering as Linux has made its inroads in company IT infrastructures, and many people have begun to question Sun's long-term viability. Short on details, Sun's announcement indicates that it plans to directly confront the trend to open source and make itself relevant in the open source world. Solaris is a robust, scalable, and widely used operating system. Making it available under open source could blunt Linux's progress and help Sun flourish in the future. The challenge for Sun will be to craft a strategy that makes useful software available under open source license while keeping enough software commercial to allow a significant revenue stream.
Computer Associates has made its relational database product, Ingres, available as open source. Because Ingres probably does not provide a significant revenue stream to CA, the switch to an open source license does not threaten CA's corporate viability. However, CA believes that it can make Ingres a significant resource in IT infrastructures and thereby provide an opportunity to sell additional CA products and services to Ingres open source users.
These are probably only the first examples of what will become an ongoing trend. As products age or lose the contest to become corporate standards, their providers will seek to maintain or even enhance their viability by migrating them to open source availability. The challenge for IT organizations considering using open source versions of existing commercial products is to assure themselves that the sponsors are really developing an open source community-based product, not just fobbing a tired product onto someone else. Open source works when there is a large and active user community involved in developing and implementing a product; committing to an open source product that does not have a good community base is like buying a Yugo - not a good decision no matter what the price. However, if the provider is truly committed to making an existing commercial product a viable open source product, the benefits to users might be enormous because they can realize the benefits of past product investment. Careful analysis will be required to ensure that IT organizations successfully take advantage of this trend.
As a side note, CA also made two other open source announcements at the same time they disclosed their Ingres plans. CA is planning to marry JBoss support to the Ingres product, offering IT organizations a complete CA-supported open source transaction and storage infra-structure. Even more interesting, CA also announced its plans to support Zope and to participate in the creation of the Plone foundation. Zope is a fairly well-known open source content management system, while Plone is an open source portal product built on top of Zope. CA's Zope and Plone endorsements offer further evidence of the trend to use open source as a business model. Zope and Plone are likely to become much more widely known and used as a result of CA's endorsement. This announcement indicates that less-mature open source products are gaining support from established technology providers.
Open Source Models Are Adapted by Project Avalanche
Project Avalanche is a cooperative effort started by a number of large IT users who were frustrated by the sunk costs in IT applications developed as one-off projects. Project Avalanche is an effort to develop and distribute these applications on a shared basis. Project members include Cargill, Best Buy, Medtronic, Imation, and Thomson.
Companies join Avalanche for $30,000. Members may, at no additional cost, use software donated to the cooperative by other members. Avalanche provides the following benefits to its members:
Project Avalanche represents an effort to amortize the cost of common and important IT functionality across a large user base. It explicitly addresses questions of intellectual property ownership to assuage concerns of SCO-like disputes through its control procedures. Project Avalanche uses quasi-open source techniques to deliver value to its members.
While a very positive development, Project Avalanche does not take as full advantage of open source techniques as it could. One of the most powerful aspects of open source software is the value created by a large user base, made possible by free distribution. Project Avalanche conflates legal protection with access to donated software. A better approach would be to allow anyone to use the software and contribute modifications at no cost, while offering legal protection for a fee. The larger community would deliver better products to all participants, while only those members who are concerned with intellectual property issues would need to invest in the cooperative's protection. Nevertheless, Project Avalanche is a further morphing of open source and provides yet another alternative for IT organizations to gain control and reduce costs.
The Bottom Line
All five of these trends reflect the growth and evolution of open source. The drive for control and cost effectiveness will impact every IT organization; it is likely that within two years every software discussion will include the question "Is there an open source alternative?"
What these trends indicate, however, is that further open source initiatives are unlikely to follow the Linux model: open source made palatable by a broad coalition of large technology vendors. IT organizations will need to become more self-reliant and make their choices based on their own analysis. IT organizations will need new tools to help them in their open source efforts. Please see the accompanying sidebar for one such tool, the Open Source Maturity Model.
The Open Source Maturity Model
It's easy to overlook how sophisticated IT organizations are in dealing with commercial software vendors. They tend to be very capable of creating RFPs, defining SLAs, negotiating contracts, and so on. Furthermore, they have honed their processes for implementing and supporting commercial software products.
However, as the accompanying article points out, the world of open source is evolving rapidly, with many different development and distribution models. The much smaller margins for open source products dictate that other organizations deliver product elements most IT organizations expect the software supplier to provide. For example, few open source projects create or deliver training. If you want training, you need to find an alternative source. Processes developed to cope with commercial software products won't work with open source products. What should an IT organization do?
The Open Source Maturity Model (OSMM), presented in Succeeding with Open Source, was developed to assist organizations with this challenge. The OSMM defines a three-phase process for selecting, assessing, and implementing open source software.
Using the key software concept of maturity (i.e., how far along a product is in the software life cycle, which dictates what type of use may be made of the product), the OSMM assesses the maturity level of all key product elements (software, support, documentation, training, product integrations, and professional services). The output of an OSMM assessment is a numeric score between 0 and 100 that may be compared against recommended levels for different purposes, which vary according to whether an organization is an early adopter or a pragmatic user of IT. Table 1 lists recommended minimum OSMM scores.
In Phase 1 of the OSMM, an organization runs through each element with a four-step process: define requirements, locate resources, assess element maturity, and assign element score. Based upon the organization's particular requirements, the available resources are assessed for their maturity and a score between 1 and 10 is assigned. The output of Phase 1 is a set of scores for each of the key product elements.
Of course, not every product element is of equal importance. Software is fundamental; support is critical; documentation, though necessary, is less important than the previous two elements. In Phase 2 of the OSMM, weightings are applied to the individual element scores to reflect their overall importance for the product maturity. Default weightings are recommended, but each organization is free to adjust the default weightings to reflect its particular needs. Table 2 lists the default weightings of the OSMM.
An overall product maturity score is calculated in Phase 3 of the OSMM. This can be compared to the recommended minimum scores to determine whether the product is suitable for an organization's needs. The score can also be evaluated to determine if there are problems with the product that the organization needs to mitigate. The recommended minimum scores are, of course, just that: recommendations. The organization does not have to follow them rigidly; they are intended as guidelines to help it determine if an open source product will serve its needs.
Reader Feedback: Page 1 of 1
Latest Cloud Developer Stories
Subscribe to the World's Most Powerful Newsletters
Subscribe to Our Rss Feeds & Get Your SYS-CON News Live!
SYS-CON Featured Whitepapers
Most Read This Week