See post 1 of 2 here.

Recently Shaun Connolly wrote:

“I admit I have traded barbs with Savio in the past. I still like my “Open Source Community and Barack Obama” blog that got him off his uninformed rants of JBoss not being “open source” enough.

Savio appears to be stuck in a 3 month rut yet again…it’s the same old flawed analysis rehashed and restated in slightly different contexts. Comparing apples and oranges.”

As I mentioned in my previous post explaining my views on OSS, I stopped caring about whether JBoss was “truly open source”. I will freely admit that Shaun’s post helped me realize that I was being too much of a purist, without considering that JBoss was helping all OSS vendors with its success. I hope that I can repay the favor to Shaun with this post ;-)

Shaun thinks that my distinction between selling a support subscription and selling products is pointless. Matt has argued as much, and believes that there is no reason to sell proprietary (OSS or OSS-based) products alongside freely available OSS products.

Here’s why I disagree.

Shaun and Matt reiterate the dogma that the OSS business model is better than sliced bread. It is, to a point. However, in the software market, selling products drives more revenue than selling support. We’ve trained IT buyers to value software support and maintenance at ~15% of the initial price of the software product. Why do OSS proponents ignore this point?

What if Red Hat, JBoss, MySQL, SugarCRM, MuleSource, Hyperic, Alfresco and Zend all woke up tomorrow and told the market they are going to offer great OSS products and proprietary/gated access to other (OSS or OSS-based) products. The reason for doing so, as these vendors would explain, is to capture a higher percentage of revenue from the user base who is receiving a lot of value from OSS today. The additional revenue would be used to fund further product development, both for the OSS product and for the proprietary/gated access product. The extra revenue would be invested in growing the company, thereby expanding the reach of OSS in general. Now, OSS purists would claim that these vendors are trampling on the true vision of OSS. But, wouldn’t the average user benefit from better products that reached a wider user base?

Yes, some of the above mentioned OSS vendors already sell products under the guise of a service. By doing so, these vendors minimize the value of the product itself. Worse, they make it next to impossible for an OSS startup to follow a business model other than one that will severely limit their revenue size in the future (i.e. selling support subscriptions). They make it taboo to sell products, because, you know, “a real OSS vendor wouldn’t need to stoop to doing that”. OSS proponents tend to err on the side of purity when the markets that we (well, you, I work for a commercial vendor) participate in are based in pragmatic reality. As I realized when I questioned whether JBoss was “truly open”, it doesn’t matter as much as whether a large number of users benefit and the vendor is open ‘enough’.

I am aware of one well known OSS vendor who will be announcing a gated access product in addition to their lineup of ‘pure’ OSS products. MySQL has gone down this path with MySQL Workbench. I cheer for them and other vendors wise enough to learn from commercial vendors and brave enough to act in the face of OSS purity.

You can disagree with my argument for selling proprietary OSS products (heck, I could be wrong!). But I ask that you do so after considering the historical success of selling software products versus the predicted success of selling support, the latter of which is to represent 1.8% of the 2011 software market.