Despite hefty successes like Hadoop, Linux, and Android, the open source movement has come under intense pressure recently. On the one hand, the New York Times' Quentin Hardy questions "whether it makes sense to build free stuff at all" given open source's commercial floundering. On the other hand, InfoWorld's own Galen Gruman chides "open source mobile efforts" as having "a history of failure."
In these and other broadsides, the critics miss the open source forest for the trees.
Redefining open source
For one thing, neither critic seems to have their facts straight about open source success. On his way to (rightly) ripping apart Moblin, Maemo, MeeGo, and Tizen as open source mobile failures, Gruman conveniently overlooks Android, the world's top mobile OS.
Oh, Gruman mentions Android, but he insists "Android is not a really an open source development effort," because "Google does the development, not the open source community."
Well, welcome to the modern world of open source, Mr. Gruman, where the vast majority of successful open source projects are exactly like this. Take a look at OpenStack, Linux, or any other commercially mainstream open source project. They're all written by (wait for it!) companies -- not for peace, love, and freedom, but for sales, market share, and customers.
Even if we allow that such projects are still fueled by a community of commercially motivated companies, plenty of projects are heavily guided by a single company: MySQL, MongoDB, Drupal, Alfresco, and many, many more. Are these not open source simply because they don't fit the mythical (and mostly false) conception of open source, which is a merry band of hippies writing software for fun?
Ridiculous -- open source has never fit that mold.
Making money with open source
Then there's Hardy's contention that open source is a commercial flop. He argues:
The open source method may be effective if enough people play along, but it does not make money in itself. Moreover, by definition it implies that open source projects have many more mistakes, bad code and failed efforts on their way to succeeding, compared with conventional projects.
The first point is a trite truism, and the second point is patently false.
On the matter of code quality, there are plenty of studies (here's Coverity's) that show open source software quality is often better than that of proprietary software. The reason isn't some magical formula but common sense: If you're going to allow others to see your code, you're more likely to ensure that code is actually worth looking at.
On the first, it's true that directly monetizing open source code is hard. But this has never been the most successful commercial model for open source. Most of the world's top open source contributors -- such as Facebook, Google, Twitter, LinkedIn, Netflix, and the rest -- sell services on top of open source, rather than selling software.
Even companies that look like the traditional "open source company" follow this model, selling software and services that complement the open source code. By my estimates, Cloudera is doing in excess of $100 million in sales, with other open source companies like Acquia, Hortonworks, MongoDB (my employer), and DataStax faring very well.
Don't get me wrong: There are plenty of downsides to open-sourcing one's software. For example, doing so makes it harder to sell that same software.
There is no other way
But here's a reality check: Open source is now simply the way we write software, as I argued recently. Or as Cloudera co-founder Mike Olson notes, open source is increasingly mandatory:
It's pretty hard to build a successful, stand-alone open source company. Notably, no support- or services-only business model has ever made the cut ....
Separately but simultaneously, there's been a stunning and irreversible trend in enterprise infrastructure. If you're operating a data center, you're almost certainly using an open source operating system, database, middleware, and other plumbing. No dominant platform-level software infrastructure has emerged in the last 10 years in closed source, proprietary form, despite huge investment and tremendous efforts by traditional closed source vendors to hold back the IP tide.
Olson concludes, "You can no longer win with a closed source platform, and you can't build a successful stand-alone company purely on open source." In fact, some would argue you can't even learn to code anymore without relying on open source.
Gruman and Hardy seem to want to suggest that we have the option to go back to the good old days of proprietary software, when revenue was easy. Meanwhile, the poster children for this model -- Oracle, IBM, and the rest -- have been missing more quarters than they've been hitting lately as they grapple with open source-fueled industry trends like cloud and big data.
Open source isn't perfect, either as a development or as a commercial model. But echoing Winston Churchill, it may well be that open source is the worst form of software, except for all those other forms that have been tried.
This story, "Fresh attacks on open source miss the mark" was originally published by InfoWorld.