In the software field, the major standardization processes are based on the RFP (request for proposals) model, which entails the selection and adoption of specifications based on implementations that already exist or are at an advanced stage of development. These models have proved in time to be more effective than processes based on work committees commended with the mission of defining a specification without having a reference implementation. Any company’s or organization’s ability to influence the specifications of any standard can today be said to be subject to it being able to justify that there are reference implementations to support its proposals.

Even if the ongoing standardization processes are sound, experience has shown that not all standards are ultimately consolidated. What’s more, the risk of adopting standards that do not have enough broad-based support from industry has been high in the past. For example, consider the high price paid by organizations that, from 1993 to 1995, staked their bets on OSF/DCE as standard middleware for developing distributed systems or, from 1994 to 1998, backed the adoption of standards like ODMG-93 (object-oriented database management standard) or CMIP (network management standard). This risk is heightened by the fact that traditional software components suppliers (ISVs — independent software vendors) often withdraw their support for any standard unless they have a solid position as suppliers in face of competition, forming, if necessary, alliances with other suppliers in the same circumstances to create alternative standards. This continuous skirmishing among suppliers concerning under-consolidated platform areas, known as the “standards battle”, has meant that neither the technical soundness of a new standard nor user support have been enough to guarantee its long-term support by industry to date. On the other hand, waiting until a standard is well-established to assure risk-free adoption is not a good solution either, as it means surrendering the possibility of exploiting the differential advantages of a new technology over its predecessors.

Open source software is taking shape as an effective instrument for speeding up the definition of any software standard. The availability of the source code of a standard specifications implementation does away with any ambiguity there could be surrounding the interpretation of those specifications. This speeds up the standardization process enormously. Organizations like the European Union’s IDA Open Source Software Observatory have explicitly recognized open source software’s impact on this point:

“Open Source Software (OSS) tends to use and help define open standards and publicly available specifications. OSS products are, by their nature, publicly available specifications, and the availability of their source code promotes open, democratic debate around the specifications, making them both more robust and interoperable.”

Also, the availability of open source reference implementations is proving to be an effective tool for consolidating software standards. On the one hand, participation in open source software projects is open to any agent, and other members of the community can pick up the baton from anyone who has participated in and decides to drop out of the development. This is possible because the source code is available. As long as the technology is sound and appeals to a big enough number of talented developers or companies that perceive potential business opportunities, there are good guarantees of continuity. Note here that initial investment, a barrier for any company intending to get involved in evolving the open source reference implementation of any standard, is low (really, the absolute minimum) and is related to the effort needed to get acquainted with the developed software. On the other hand, when the open source reference implementation is associated with a permissive licence (like Apache) allowing the manufacture of closed-source products, there is a much better chance of there being enough products implementing the standard. This is because the barrier of initial investment required to develop the product is much lower: other parties have defrayed the costs of developing the part of the software reused to build the product (the part about which the manufacturer makes no claim to differentiation through a proprietary development) and it can be used absolutely free of charge.

MyMobileWeb is an example of a project within the Morfeo community where we are experimenting with these principles. The goal of this project is to boost the definition of standards that will be vital to materializing the Mobility Web concept. Open source reference implementations of W3C standards are being developed as part of this project. The aim is to speed up and consolidate standard development. The partners involved in the development of MyMobileWeb project technologies (Telefónica I+D) consider that the development and standardization of technologies that bring about an explosion of mobility accessible web sites will lead to a sizeable profits from indirect income (higher mobile telephony demand).

Creating new opportunities for integrated solutions

It is not easy to launch new products in market segments already occupied by major software manufacturers following the traditional model for commercializing software. This model focuses on product packaging, product distribution via marketing channels and income generated by use licences and maintenance fees. The investment in marketing activities to publicize the product, product packaging activities or the development of support and distribution channels to rival competitors’ that this model calls for is prohibitive. Use licence prices have to be set at high levels to get a ROI within an acceptable term, and this is likely to dowse any interest there might be in taking the risk of trying out a new product.

The open source software marketing model breaks traditional marketing model rules and opens up opportunities for any newcomer to software product manufacturing.

• The newcomer “releases” its product under an open source licence, sweeping away the entry barriers as regards use by market segments with lower purchasing power (private persons, SMEs) thanks to its price (it is free).
• No marketing effort is required: if it is a good quality product, the users themselves will publicize its good points (network effect). It should be stressed here that the market segments with lower purchasing power that first adopt the product and publicize its strengths may be segments that the manufacturer would never have targeted (even adopting a traditional software marketing model). Therefore, the strategy does not necessarily entail any loss of clients.
• No distribution channels need to be developed: the network is the channel (the software is downloaded from the Internet).
• When the software starts to make a name for itself, the big companies will start to take an interest in adopting the product. However, they will want it to be packaged with services (support, maintenance, consultancy) provided by someone with proper credentials. It is with the sale of services like these direct to the client or to third parties working for the client that the manufacturer will perceive the income to first cover the manufacturing costs and then earn profits.
• Open licences do not stop a product from also being distributed subject to a commercial use licence under which you pay a fee for use, for example, of more advanced versions or to lift some restrictions on use.

Note that the newcomer may be a single company or a group of companies that opt to share the product development and maintenance costs, where the business opportunities focus on income derived from offering services, as allies or competitors, to customers that are members of their respective customer portfolios.

The magnitude and projection of open source software is unquestionable. Nevertheless, these are software products in whose development Spanish companies or organizations have not participated or are not participating. North American companies, like RedHat, IBM or HP, especially are seizing more of the business opportunities materializing from selling services based on open source products (support services, consultancy, solutions integration), and, as members of the European Commission have repeatedly mentioned⁴, this is happening all over Europe as well as in Spain. These companies actively participate in the development communities associated with open source software on which basis they then develop their services offer. They are then able to cash in on the above competitive advantages that they gain from this active participation.

The absence of open source products related to basic software technologies still at an immature standardization stage (technologies associated with Web Services, with the concept of the Semantic Web, BPM and workflow technologies, etc.) or regarding business process support platforms where there are no alternatives to commercial products (products like Siebel for CRM solutions, SAP for ERP solutions, JD Ewards for supply chain management solutions, MDSI for workforce management, etc.) is an opportunity there for the taking. To take advantage of this opportunity, Spanish companies, universities and research centres should sponsor the establishment of and participate actively in an open software community working on unexploited areas like these. Such a community would be instrumental in the development of what is today a non-existent Spanish software technology and product manufacturing industry. The companies actively involved in the community would benefit from the same sort of competitive advantages that companies like IBM or RedHat are exploiting as a result of their participation in the development of open source software products.

The companies involved in developing standard reference implementations as per an open source model have the credentials they need to actively participate in the respective standardization processes and to develop a services offer based on these implementations (or equivalent software products) with maximum guarantees for customers. These credentials are the factor that differentiates these companies from competitors that have not participated in reference implementations development. Companies like IBM are adopting this strategy and are becoming very successful in the growing business of services for open source products.

Note that the development of business opportunities around the services offer for technologies developed under an open source formula are not confined to the companies directly involved in the development of these technologies. There is a potentially promising formula involving a win-win relation between the companies developing the technology and SMEs that would play the role within these communities of selling but not directly developing products. Under this model:

• The SME acts as a sales channel, working with the technology manufacturing companies to exploit the opportunities for selling support and maintenance, consultancy, solutions integration or training services to their customer portfolio. In this case, the SME and the manufacturing companies will design schemes to distribute income from the sale of those services.
• In return, the SME manages to add to the portfolio of services that it can offer based on open source platforms backed by established alliances with technology manufacturers.

Again, some of the projects within the Morfeo community are being used as a test bench to find out whether these premises are sound. Prominent examples of this are projects related to the MyMobileWeb platforms, aimed at improving the construction of web applications and gateways accessible from mobile devices, and SMARTFLOW, aimed at improving the construction of dossier/task management systems. In both cases, a growing number of SMEs are signing up to commercialize the developed technology in conjunction with the companies behind the technology development.

Boosting R&D&I activities

The formation of an open source software community to serve as a framework for developing R&D&I projects has sizeable benefits. First, the projects that are carried out in the community are run “live”. They are, therefore, a forum where the community members can easily demonstrate in which research areas they would like to work with other agents, like sector companies and public research organizations.

Second, the design of projects within an open source software community is an incentive for leading university research groups to get involved in R&D&I projects of interest to industry. This is because the model promotes the attainment of goals that are of unquestionable importance to any research group, like the recognition of intellectual property rights to the work they do and entitlement to disseminate the results. As regards the first point, remember that software developed according to an open source formula usually recognizes the intellectual property rights to the work done by any project contributor, although provision may also be made for formulas enabling the original product manufacturer to exploit dual licensing models. As regards the second point, a research group will find it appealing to have a platform (community web site for the project) by means of which to disseminate its work bearing an applicability seal awarded by firms, especially if these companies have a strong brand position (as is the case of Telefónica).

Third, the adoption of an open source software development model is a way of speeding up the release of final products derived from R&D&I projects run jointly by more than one agent. Joint R&D&I projects addressing software development potentially for manufacturing products on which to build a commercial offer have been based to date on two key models:

• Some projects do not consider the development of a final software product that the project participants can sell at all. The developments undertaken in the project are for test of concept or experimentation and are far removed from what is a final product: the focus is on researching aspects related to the definition of specifications, architecture, how to integrate technologies, etc. Each company is entitled to use the project results as a basis for building a product, but will do this after the fact, outside the consortium. This model prevents management problems regarding intellectual property and the rights of use, modification or distribution of the developed software. On the other hand, it delays the release of saleable products.
• Other projects do consider the development of source code that can be integrated into saleable products. However, it is vital in these projects to establish a contractual framework among the consortium members that specifies from the very start how points like intellectual property or the rights of use, modification and distribution are going to be managed. The biggest drawback in this case is the difficulty of putting together agreements like this. This can lead to a delay in the project’s kick-off and, consequently, product release, if not a deadlock preventing the project from ever getting off the ground.

The open source software development model introduces another alternative. The projects can cover the development of saleable open source software within the framework of the project based on a clear and well-defined management of intellectual property and the rights of use, modification and distribution that does not need to be negotiated because it is already there: it is stipulated by the adopted open source code licence (GNU/GPL). The only condition is that the companies participating in the project must accept that the business opportunities derived from exploiting the project results will stem from aspects (additional services, sale of hardware or equipment on which the software is installed, etc.) other than income earned through payment for use of the developed products. An additional benefit is that this model also encourages new project participants to sign up.

The Morfeo open source software community aims to test the soundness of these premises by acting as an incubator of R&D&I projects and experimenting with a win-win relation model between companies and research groups (based at universities and technology centres). Several projects have been successfully incubated within this community, certifying the points made in this section as valid. However, more headway needs to be made along these lines.

Worthy of note is the fact that the Morfeo community has been selected as one of the nodes that are to form part of the pan-European network of open source software development centres of excellence defined in IST FP6 IP Qualipso. The ObjectWeb community, an important reference point in the field of Java J2EE application servers, will also be part of this same network. This pan-European network could become a “virtual factory” developing reference implementations of technologies and products to boost the creation, use and access of tomorrow’s electronic servers. On this point, one of Morfeo’s key goals is for many of the competencies within the pan-European Qualipso network to be located within the Morfeo community.

Conclusions

Electing to develop open source software does not mean that you are against developing software to be sold under a use licence. Both models are fully legitimate and valid.

Developing software according to an open source model simply means exploring new ways of doing things. These new ways can materialize as new opportunities that some companies or organizations can capitalize upon.

Morfeo is an innovative project as part of which companies, universities and other organizations are exploring what new opportunities developing open source software can offer.

References

1 http://www.dwheeler.com/oss_fs_why.html

2 Libro Blanco del Software Libre en España 2004, available at: http://www.libroblanco.com/document/255-2004.pdf

3 “Predicts 2005: Open-Source Software Proliferates” Gartner Group Report

4 Talk by Jesús Villasante, Holland Open Software Conference 2005. Brief summary and link to the article published by Computer Business Review. OnLine available at: http://istresults.cordis.lu/index.cfm/section/news/tpl/news/Briefs/briefs/ID/76994

Pages: 1 2 3