A blog for the enterprise open source community
The future of open source licensingMatthew Aslett, August 25, 2010 @ 7:57 am ET
There has been plenty of discussion around our recent post arguing that we have entered a new era of open source commercialisation strategies.
“Open source 4.0″, characterized by corporate-dominated development communities, is upon us with increased emphasis on collaborative development for non-differentiating code. In comparison the previous era was dominated by vendors that eschewed the potential advantages of collaborative development in favour of control over the future development of the project.
The shift has occurred, as Matt Asay notes, “because the world has sorted out where competitive advantage lies (e.g., data), and is therefore willing to collaborate on non-differentiating code while keeping everything else firmly proprietary”.
We noted in our previous post that the participants in these corporate-dominated development communities are not necessarily vendors and that we are seeing increased involvement of companies that we would traditionally think of as consumers of software, rather than contributors.
However, it is worth noting that a significant proportion of the vendors that we see involved in these communities (and certainly those that we expect to benefit most commercially from lowered R&D and indirect revenue opportunities) are traditional closed-source vendors.
The shift has also occurred at a time when we see a surge in opposition to open core strategies in which the dominant vendor makes closed-source extensions available to paying customers. As Matt explains, this has led to a somewhat surreal scenario in which opponents of open core are celebrating a shift towards a model that favours the traditional closed-source software vendors.
The situation has also given cause for concern, with Glyn Moody highlighting that the Apache licence is being widely adopted for new projects and that “anyone can take the code and turn it into a proprietary offering.”
Glyn argues that this “is worse than the GNU GPL with copyright assignment”, where a single copyright holder is able to provide a closed-source version. Is it really worse to have a situation in which everyone has an equal opportunity to go closed-source than one in which the control and power lies with a single vendor?
Either way, Glyn also notes that the ability of the copyright holder to act as a monopolist is “hardly what Richard Stallman had in mind when he drew up the GPL”. This is probably true, but it should be noted that Stallman has defended selling license exceptions (also known as dual licensing), which is the practice that enables vendor-controlled open core strategies (which Stallman opposes).
In fact, it is worth considering that the issues that seem to cause the most controversy around open source-related business strategies – vendor-controlled open source projects, open core licensing, copyright assignment, and dual licensing – are all perpetuated by copyleft and the GNU GPL.
It doesn’t have to be the case that the GNU GPL leads to a dominant open source vendor, of course. Glyn explains how assigning copyright to a non-commercial entity, such as the Free Software Foundation, avoid this problem. Another approach (although one that has problems of its own), is to ensure that individual contributors own the copyright to their own contributions.
Glyn makes a case for the GPL combined with copyright assignment to a non-commercial entity being the preferred option for corporate-dominated development communities, arguing that “it produces a completely level playing-field for companies. Anyone can take that GPL’d or LGPL’d code and use it as the basis of a commercial offering; however, no company can gain an advantage over any other by offering a closed-source version too.”
It is interesting that offering a closed-source version is seen as providing an advantage. I would argue that a situation in which everyone is able to offer a closed-source version creates the same level playing-field. While it is perfectly possible for a vendor to take without giving, doing so negates the benefit they would gain from being part of an ongoing collaborative development process.
From a commercial perspective, the incentive is to share non-differentiating code. In comparison there is little or no incentive for vendors to create non-commercial entities for copyright assignment. The ability to create closed-source versions, and/or offer complementary closed-source software and services, is one of the main commercial drivers for vendors taking the corporate-dominated development community approach.
Non-vendors may well have an incentive for preventing commercialisation of their efforts, but while we see increased involvement of non-vendors, it is still likely to be traditional closed-source vendors that dominate these communities.
The GPL is certainly not going away any time soon (any more than we have previously discussed) but commercial incentives mean that non- and weak-copyleft licenses are highly likely to be the licenses of choice for the projects that define open source 4.0.
UPDATE – I am, of course, generalizing here. Carlo Daffara provides an excellent analysis of the factors that influence license choice, including existing code requirements and intellectual property. – UPDATE
We explain more about our theory of the evolution of commercial open source in Control and Community, the follow-up to our Open Source is Not a Business Model report, which is now available. The report provides more context for the economic motivators and issues involved in the various models, as well as updated research on which vendors are following which strategies, and why, as well as a survey of 286 open source software users to uncover what they make of it all. The report will be freely available to CAOS subscribers. For more details of the CAOS research practice, and to apply for trial access, click here.
Comments (4) Categories: Business strategies,Licensing,Software