2005 was a great year for open source developers and solutions, with a dramatic boost in credibility, tools and respect. As part of this, there are an increasing number of companies commercializing open source: adding value through installers, packaging, coordinated releases, technical support, management utilities and formal product planning.
This has the feeling, though, of domesticating the wild spirit of open source and turning it into another IPO-driven, VC-backed, competitively focused economic model. I’m reminded of Joni Mitchell’s verse from “Big Yellow Taxi”:
“They took all the trees
And put them in a tree museum
And they charge the people
A dollar and a half just to see ‘em”
The rush to create for-profit companies built on open source assets may create a new generation of forest rangers protecting our open source resources, or we may instead see clear-cutting of the values at the core of the open source movement.
First, Some Background
Open source has been around for a long time, with LINUX started in 1991 and the remainder of the LAMP coalition of similar age (Apache in 1995, MySQL in 1995, Perl in 1987, Python in 1991, PHP in 1995). These have a huge developer community and impact on the commercial computing world. Notably, each has had at least a decade of organic (non-commercial) growth, enough to survive and stay vital as for-profit players emerge. [Red Hat and SUSE with LINUX distributions; Zend and ExpressionEngines in PHP, MySQL AB with support and training.]
There’s a newer group of companies commercializing narrower open source solutions. SugarCRM‘s customer management system is open-sourced to allow organizations to extend their solution. XenSource extended the Xen project with tools for system virtualization. Groundwork is building on the pre-existing Nagios network monitoring application. In these cases, companies are starting from younger open source projects (or creating new ones from scratch).
I’m assuming that every existing open source project is under some entrepreneur’s microscope (and on a few VC funding maps).
Good or Bad?
Yosemite was a pristine area through the 1860s. Making it a national park made it much more accessible to the general public. Park rangers, roads, campsites and hotels, bathrooms and bear monitoring let hundreds of thousands of visitors see the park — and experience the great outdoors. Of course, these “great outdoors” are compromised as they are domesticated… but smaller numbers of rugged individualists can still travel elsewhere for more authentic, unspoiled interactions with nature.
By analogy, the open source experience is becoming tamer. Less expert users (including corporate users with budgets!) can get commercial-quality software that’s installable in hours, not days. Downloads and installers work predictably, applications are more complete, professional-grade interface tools are available, and you can pay a knowledgeable support engineer to help you get things right. This is progress for the 95% of us who are intimidated by make, awk, and SourceForge.
Companies want to be paid for their commercialization, though. Dual-licensing — with GPL and conventional licensing available for the same software — is becoming more common. jBoss, for instance, is starting to supplement its “free software, paid support” model with some components only available to paying customers. We’ll see an increasing number of open source companies charging for important add-on or management software, which historically would have been fully open sourced.
Corporate folks (like me) may apply eco-friendly labels to themselves and the commercialization process: shepherds, park rangers, managers, protectors of the pioneer spirit. Experienced open source “Earth First-ers” may instead think of strip mining, clear cutting, or mass capture of wild Mustang herds. Regardless of the name, this stuff coming to an open source project near you.
The scarce natural resource for all of this, though, is the open community of volunteer developers who make bring these projects to life. Without enthusiastic unpaid software engineers reviewing and contributing code, open source efforts look just like traditional commercial software development. Christmas tree farms.
What happens next?
With the big push to commercialize open source and tame some of this wilderness, here’s what I think will happen next:
- Fewer volunteers. Companies are recruiting the best open source engineers as formal employees, just as many more open source projects are sprouting. The pool of informal, unpaid, community-minded developers will shrink. In its place, we’re seeing paid developers at large companies (IBM and HP) who are assigned to work on strategic open source efforts as an alternative to in-house application development. This signals a rapid Darwinian shift where technical resources depend increasingly on corporate product roadmaps and business development.
- Over-marketing to the community of developers. The not-so-secret ingredient for open source is an ecosystem of volunteer developers. Marketers will be asked to deliver and measure outreach to the open source community. Watch for overly aggressive promotions and inducements that eventually resemble soft drink advertising.
- “Instant” ecosystems. As many of the strongest open source projects are co-opted by for-profit ventures, companies will try to create new communities from scratch. They may donate existing product lines to newly-formed non-profit organizations or for strategic reasons (unseating market leaders, shifting licensing models). This will look strikingly similar to earlier private ecosystems: commercial software companies evangelizing their proprietary architectures to outside developers in order to build market momentum and lock in support for strategic initiatives.
Together, these will make growing new open source projects more challenging. Supply of a key natural resource (volunteer developers) may not meet demand.
The open source movement is robust, but commercial encroachment is inevitable. Let’s expect the industry to “pave paradise, put up a parking lot.”