Open source + open data = Open cloud

It used to be taken for granted that the web was and always would be open. That assumption has increasingly come under fire as cloud computing has set up walled gardens for data and services...much as the desktop has done.

Tim O'Reilly addressed the threat of closed clouds and closed mobile devices to access the web in his keynote at the O'Reilly Open Source Convention.

Tim seemed to have lost interest in open source over the past few years, his interest instead turning to Web 2.0 (though he continued to recognize the need for an upgrade to the way we think of open source in terms of licenses instead of services). But somehow, somewhere, Tim re-discovered the importance of open source, this time in keeping Web 2.0 from turning into Manacle 2.0.

I'm not sure that you ever truly left, Tim, but this call to arms is timely and welcome. In his keynote, Tim said:

One of the things that's happening in cloud computing and Web 2.0 is centralization. We need to think about what will make this web platform of the future open.

So Tim articulated, and so I believe. Open source serves as a safeguard on our best intentions. I still believe licenses matter, though I concede Tim's point that they may not matter nearly as much as open services guarantees.

In the Software 1.0 world, the customer is captive to the vendor. Take Oracle and SAP, for example. Customers have bought into the software and if Oracle and SAP decide to, say, raise prices 20 percent, the customer has to go along with it. Heck, if these companies really wanted to push the envelope, they could probably raise prices 100 percent. If you're $100 million into an SAP license, what are you going to do? Complain?

In a similar but different manner, baby cloud services (SaaS) like Salesforce.com are a one-way freedom ticket: You're free to enter (and the cost of entry is very low), but forget about leaving for free. Your data sits on Salesforce's servers. It has not interest in helping you to facilitate an exit.

The same holds true for Google, Yahoo!, and other Web 2.0 companies. The clouds they are creating are closed, though Google has been doing some of the right things with open data pledges and increasing commitments to open source. (Not everyone is convinced.)

So, users need to be vigilant about keeping themselves from getting entangled with cloud lock-in. Demanding open source and open data guarantees are two ways. Requiring that the devices we use to access the cloud remain open is another way, and one that Tim suggested in his keynote. Preventing the web world from being consolidated into the hands of a very few players is another way, as Larry Lessig has argued.

Unless we want another few decades of lock-in, this time in the cloud and created by Google, Yahoo!, Microsoft (if it ever figures out that it needs to start with an open platform and then create a veneer of proprietary data lock-in), Amazon, and others, we need to wake up and figure out how to make open source apply more effectively to the web. It was open once. It can be again. But it won't be easy.