The fuss of the day: Terracotta is the new guy jumping on the Open Source bandwagon. Nice stuff, useful indeed and something I’m looking forward to use in our projects, but there is a catch. The license they adopted is the somewhat disputed MPL+attribution variant, which requires to display a prominent message on any “user interface screen”, something that seems to be highly popular among the next generation of Open Source product vendors. Waiting for the dust to settle, and leaving to the OSI the difficult issue of sorting out whether the amended license can still be considered Open Source, I will just dump a few comments before leaving for a few days out enjoying the Tuscany landscape.
As an Open Source guy committed to community based development, and as a strong believer of commons-based peer production, all I can say is that I have mixed feelings about the next wave of Open Source products. I’m seriously convinced that if you take cooperation away from the picture, all you’re left with is a (potentially useful) piece of code, somewhat without a price tag. Not bad, but not a big deal after all. I sincerely hope someday these guys will see the light and move from their current defensive approach, based on a strong quid pro quo paradigm and a somewhat irrational fear of the bogeyman making money with their software, to a full-fledged open strategy, built upon gathering people together to build better software, optimise the development process and fully leverage the power of Open Source development, while building a competitive advantage based on reputation, positioning and branding.
As utopian or socialist as this may sound, the theory behind community-based development is factual, pragmatic and capitalistic: there is little value in software developed behind firewalls, no matter how it gets distributed. The MPL+attribution license makes Open Source licensing nearly worthless, unless you’re fine having your application potentially look like a Times Square billboard: the paid-for license is the price you have to pay to make advertising go away. Let’s call a spade a spade: attribution, as phrased in the BSD and Apache license, is unobtrusive and meant to both give credit where credit is due while informing the final user about the origins of software. This new licensing model is not about attribution, it’s about advertising as a way to pay for software, somewhat the server-side version of shareware pleads for money: it’s about crippling the open source version, making it nearly useless. What we have, by all means, is a license which is trying to limit usage of the “free” edition, and a factual, if not literal, betrayal of the Open Source idea of increasing circulation of software.
I don’t have a problem with quid pro quo, mind you. We all have bills to pay, and all the yadda-yadda I’ll skip for brevity sake. Quid pro quo, however, is nothing but a barter which needs to level out: the more you offer, the more you get in return, and the opposite is true as well. Crippling the distribution with tricks to make people buy a commercial value means lowering the value, getting less in return and making software look like a Petri dish culture in comparison to the rain forest. Don’t expect bug fixes, don’t expect collaborative software development, don’t expect communities, don’t expect diversity to provide evolution: all you get is a ride on the Open Source train, which in my opinion won’t be able by itself to sustain business in the long run, as customers will either have an hard time understanding what Open Source really means (and run away) or develop antibodies to recognize crippled open source editions, driving to paid-for versions, from what really meet their needs (such as lock-in reduction, vendor indipendence and acquisition at the point of value, none of which is fully accomplished by the new productised Open Source approach).
Waiting for the OSI jury to return the verdict about compliance of these advertising clause with the Open Source guidelines, all I can do is stressing once more why we need an Open Development definition. No hard feelings towards SugarCRM, Alfresco, Mulesource and friends who just happen to focus on Open Source as a new software distribution rather than a disruptive new development model: it’s not a problem of mudblood vs. pureplay. I’d love those guys to embrace community-based development, and I’d be happy to start a conversation to let them know why I feel they are losing business, yet the software market is large enough to accommodate everyone. As long as we provide our customers with accurate taxonomies so that they understand what they are dealing with.