It’s kind of sad that anyone even needs to say such a thing, but apparently Matt Gemmell needs a reminder. In the middle of many good points about the Sparrow acquisition (good even though I found Rian van der Merwe‘s take even more insightful), Matt drops this huge turd.

It’d be a shitty, reduced, pale imitation of what Sparrow actually was, because the developers would have had to take so much time off to attend the funerals of their families who had died from starvation.

WTF, Matt? That’s incredibly offensive to those who’ve actually had to attend the funerals of their families, or have seen real starvation up close. Even when one strips away the hyperbole, though, it’s still quite wrong. I work on GPL software, and my family hasn’t starved. We’ve done quite well, in fact, and we’re only one among thousands of Red Hat families who could say the same. That’s because there’s a real open-source business model that Matt doesn’t seem to understand. Sure, you give up the immediate revenue from the software purchase cost, but it turns out that recurring-revenue models outperform one-time-revenue models and those are still possible with open source. People will still pay for the earliest and easiest updates. They’ll still pay for guaranteed support by dedicated support professionals instead of “whenever the developer feels like it” support. They’ll still pay for training and certification. They’ll pay more for all of those things as a package than separately, which may be a quirk of human psychology but the effect is still real. The subscription model works, and I’m part of the empirical evidence for that.

That’s all what makes open source equal to closed source, but it turns out that there are other things make it better. For one thing, even though “you have the source so you can fix it yourself” is utterly useless to 99.9% of users who lack the skills to make meaningful change, you’d be surprised how much of a difference that 0.1% can make. Even if only 0.1% of your users ever contribute anything, with thousands of users that can significantly augment your in-house development team. People who just want/need one specific feature to support their separate business, and don’t mind sharing that feature with the rest of the world, might add it themselves. Alternatively, those who lack the skills/resources do so it themselves can sponsor such development either by the original developers or by various consultancies. Take a look at gcc, for example, which has benefited from this phenomenon many times allowing hundreds of developers to make a good living despite the existence of many closed-source competitors over the years. The net result is that open source can often enable software to evolve more quickly than the same development team could do with closed source.

There’s also an ecosystem element to this. Besides accelerating development of a particular package, open source can accelerate development of all alternatives in a technical area. Open source gives newcomers a way to develop valuable domain-specific skills. Similar open source projects can share libraries and frameworks, reducing duplication of effort. Also, the mere existence of an open source competitor can increase knowledge about what’s possible, what users really want, what problems still need to be solved, etc. That’s valuable knowledge, which might otherwise take a lot of resources – or trial and error risking customer defection – to gain. Increasing the total number of users within a category, even if they’re not your users, helps the whole category – including even the closed source parts of it – make technical progress. For example, look at NoSQL or Hadoop. Much of what they do is not dissimilar to products that existed in more closed form before, but once there were a significant number of open source offerings the number of users (and developers BTW) absolutely exploded. The result has been a phenomenal increase in the pace of development across the board, for all participants.

Open source both creates a bigger pie and supports monetization of that pie. Nobody starves. The only people who don’t benefit are those who remain blind to how the open source business model actually works, who think they can use FUD to compete against it. Well, too bad for them.