How Open Source Works, or Why HPCC Will Fail

Having followed open source for nearly a decade now, I've learned a little about how the business works.

It's based on what appear to be contradictory impulses.

1. It's driven from the bottom-up. It's based on community, even if the community involved isn't giving you money or code. Without the passion and mission of users, you're dead.

2. It's entrepreneurial. No new business can grow fast without someone driven at the helm. Even if you're not about the money. The Free Software Foundation is nothing without Richard Stallman, the Linux Foundation foundered until Jim Zemlin came along.

So when I see a press release or a web site run by an open source start-up, I can now tell without looking at the code whether the project has a Clue and thus, a chance.

HPCC Systems, for instance, which just emerged from the bowels of Lexis-Nexis to “challenge” Hadoop.  No chance. Oh yes, that logo for its lead product, dubbed Thor, is cute, but no chance.

The project manager, Armando Escalante, doesn't even have the guts to quit his day job at Lexis-Nexis. This is a code dump – a company wants someone else to bring it money by contributing to the code base.

I don't care if Hadoop is “Googlesque.”  It does have entrepreneurs running it, it has a driven leader in Doug Cutting, and it already has all the key alliances in the space. If you're going to challenge, if you're to gain “second-mover” advantage, you need to bring a community, something compelling, along with some serious business smarts.

Contrast that with Antelink, which is pushing a new search engine for open source code dubbed Antepedia.

While there are solid players in this space, notably Black Duck and Palamida, with working business models, Antelink has community support as the European version of Maven, and it has real entrepreneurs behind it, whom I met in Paris last fall.

They have a product in Release Cooker  that can generate cash flow, they're young guys with passion, they have institutional backing  (as opposed to a single code source behind them).

This doesn't guarantee success, but it's all a necessary precondition. Oh, and they also have a marketing niche, app developers, which is different enough from their rivals' focus on enterprise tools that it is bound to get them a look.

See the difference?