Commercial, Professional, and Community Open Source: Resolving the Naming Confusion

As a researcher, imprecise naming bothers me. The general confusion around the terms commercial open source, professional open source, and community open source warrants closer analysis.

First my proposal, then some litmus tests, followed by a bit of history.

  • Commercial open source is software provided as open source where a single legal entity owns the rights to the software (SugarCRM, Alfresco, etc.)
  • Professional open source is software provided as open source where a dominant firm provides services around the software without actually owning it (JBoss, Spring, etc.)
  • Community open source is software provided as open source where multiple stakeholders hold the rights and no player dominates the software (Linux, Apache, etc.)

So here are some litmus tests:

  • It is commercial open source (and not professional open source), if the open source firm can sell you usage rights to the software under a proprietary license.
  • It is community open source (and not professional open source), if the services market is fluid and not dominated by one company.
  • It is professional open source, if a single firm dominates the software but where other firms can compete (mostly) fairly on services. Typically this means that the copyright is spread among multiple parties, but the professional open source firm may hold some rights like trademarks that give some some additional leverage.

Historically, as far as I know, commercial open source was first practiced by MySQL, even though the term was only invented later by SugarCRM. A core go-to-market approach is the dual-license strategy. Only commercial open source firms can apply this strategy. The term professional open source was invented by Marc Fleury of JBoss to give open source a better reputation; mostly it refers to the provision of (professional) services around some particular open source software.

Both commercial and professional open source are go-to-market approaches and core strategies of a firm’s business model. Community open source is not a business model. If employed strategically, all it does is to curtail the revenue gathered from a particular market (like operating systems). It is used to shift revenues from the curtailed market into ancillary markets; basically a share of customer’s wallet ploy for the companies dominating one or more of the ancillary markets.


This appendix was written one day after the original post. My original post had been triggered by a rather confused academic article in a high-profile journal. So I figured I should do the leg work and review how other folks have defined these terms.

  • Giampaolo Garzarelli and Roberto Galoppini distinguish corporate, voluntary, and hybrid open source project models. Corporate open source is the same as commercial open source. In my understanding, voluntary and hybrid equal first and second generation community open source. First-gen community open source was completely driven by volunteers, second-gen community open source is being driven by volunteers and firms alike. There is no notion of professional open source, though the services business model probably always figures in.
  • Michael Nolan argues that commercial and community open source exist and need to be distinguished. He compares Open Solaris with Eclipse and views the former as a commercial and the latter as a community project. I concur with his point about Eclipse not directly making IBM money but shifting the revenues to ancillary products, as stated above and argued elsewhere in more detail. I disagree, however, that only community open source is viable in the long-run and that it cannot serve commercial purposes.
  • Matt Asay also agrees that commercial and community open source are different from each other. He argues that commercial open source has a firm at its center which gives a project focus and momentum, something that may be lacking from community open source. I used to agree with this distinction, but I think the second generation community open source demonstrated by the Apache Software Foundation and the Eclipse Foundation shows that community open source can have focus and momentum right from the start.

I want to emphasize that I’m talking about using open source as a go-to-market strategy. From this perspective, ownership rights become the core distinguishing factor. As far as I can tell, commercial and community open source are becoming the accepted terms. Professional open source as a term (not a practice) may loose traction, as people simply talk about open source services firms.

10 Replies to “Commercial, Professional, and Community Open Source: Resolving the Naming Confusion”

  1. How can you ignore the term free software ? Free software can include any of the above – the difference is in the reasons the developers state for their license, and the reasons users choose it, not in whether or how anybody in particular gets paid.
    So yes you can get commercial free software (like OutKafe, SugarCRM etc.), community free software (like python, kde etc.) and professional free software (like GNU bash, JBoss etc.).
    I think the consideration is critical – and your separation incomplete. The single largest source of free software is the FSF having produced more projects than all the others combined over the past 35 odd years. The FSF is a professional software creation organization – yet it is non-commercial, in fact it’s legally registered as a charity – but your definition would make it commercial (they COULD sell you the software under a EULA – you said nothing about whether there is a snowballs hope in hell that they WOULD).
    Then you have companies like mine, which support free software, use the phrase free software and actively promotes… you guessed it… free software. Many of the examples you mention have done the same, or chosen the term FOSS to indicate that they support BOTH the ethical values of free software AND the technical merits of the open source world (I use FOSS within my company, free software when speaking in my personal capacity – I think the logic should be obvious) .
    Either way, it’s good that somebody is trying to figure out good definitions for these ideas, but I think the process is far from even coming close to covering the breath of the FOSS worlds development models.

  2. @A.J. Venter – thanks for the comments.
    My classification is strictly written from a commercial perspective—how and why open source makes strategic sense as a go-to-market approach for software companies.
    I view a classification of free and non-free as independent of my classification. As you say yourself, any of the three commercial classes I mention can be free software or not.
    Such free/non-free classification probably has two dimensions itself, along the classic “freedom to” and “freedom from” axes. However, I’m not talking about these.
    In general, moral classifications are hard to get a scientific handle on, so you won’t find me addressing them often on this professional blog.

  3. Open Source is not inherently non-commercial. In fact, a project that is not allowed to become a commercial product is not open source.

  4. @Miks: I agree that there are many dimensions, and a commercial perspective is only one.
    Some licenses, however, make commercial use harder than others. But pretty much every open source license allows at least for commercial services around the project.

  5. Hi Dirk,
    This is very interesting. I’ve been waiting to see this distinction appearing.
    There is a big difference in openness between Commercial Open Source and Professional Open Source (using your definitions):
    1/ When the company owns the product and can use a double licence, it has more leverage to get the customers paying than when you don’t own it fully. When you don’t own it you need to run a more open open source licence than GPL in order to allow all customers to make use of the software even in non open-source situations. We at XWiki use LGPL which for me represents a good balance.
    2/ When a company is commercial open source is can gradually decide to close more and reduce it’s investement on the community side, so that it can favor the commercial offerings. Now this does not necessarly happen and still allows for a GPL fork (a Community Open Source version).
    For me this second point is key. It significantly highers the potential ROI of the business and there are highly significant long term chances to see the software being slowly closed, which will lead to a fork to Community Open Source.
    So for me Commercial Open Source is not stable long term. It will end up in a fork between nearly-closed source and Community Open Source.
    Professional Open Source is more stable although it also allows for a Community Open Source fork but this will be much less likely than in the Commercial Open Source case. The market will decide wether the initial actor deserves or not to stay dominant.
    When I started my open source business, stability was an important thing for me, so I decided for LGPL and Professional Open Source. This is binding for our users/developers that want to contribute or base their work on top of XWiki, but this is also binding for our company.
    I think there should be more understanding that being commited on one Open Source strategy is important and a good thing.
    I’m very afraid to see double licence businesses turn closed-source or nearly-closed-source.

  6. Hi Ludovic, thanks for the excellent comments! They warrant a whole article to answer them properly, so let me point out just one thing here (and answer more later): There doesn’t seem to be as much business in single-component services as we were all assuming or hoping.
    According to a recent IDC survey, which Matt Asay summarizes for us, less than one percent of all open source components at those users surveyed had attendant services associated with them. Apparently, customers are not buying nearly as much as we were expecting.
    In another informal survey discussed on the Rodrigues and Urlocker blog, several CIOs made the point that they may only be willing to pay for open source support services in the first or the first two years, after which they expect to be able to handle the components themselves.
    My take is that commercial open source may turn out to be the only viable long-term option!

  7. This is about more than name confusion, it is an important interview with
    News that MySQL was to bring out a closed-source Marten Mickos, formerly CEO of MySQL, and now head of Sun’s database group, trying to explain what is going on with the efforts to close source the database. Not sure if this is the right place to post it but please take a look.

  8. Greeting for the day
    i want to clearify my doubt actually what is the defference between Proffesional and a MBA final year pursuing my mba from TamiNadu.Under ANNA UNIVERSITY.

Leave a Reply