Last week I took part in Red Hat’s Open Source Cloud Computing Forum and discussed our view on how open source and cloud computing can complement each other.

I was therefore interested to read the views of Stephen O’Grady and Matt Asay on the same issue this week, and was pleased to see they were in general agreement with out view that open source and cloud computing are potentially complementary.

The seeds of my presentation can be found in this post I wrote a year ago today discussing the idea that open source had the potential to level the playing field and provide and opportunity for multiple cloud providers to focus on differentiating themselves on service levels rather than lock-in.

As Stephen put it this week:

“Open source cloud stacks, after all, would theoretically allow private enterprises to run on top of the exact same stack as alternate public providers. Much as Linux disrupted and commoditized the platform that applications run on, thereby improving competition in the market, so too could open source cloud projects ensure an openness to the platform that host operating systems and attendant cloud fabrics.”

Matt Asay picked up the argument and suggested the cloud platform providers could use open source as an “exit mechanism” that gives users comfort that they are not being locked-in.

During our presentation last week I argued that one of the lessons cloud providers have learned from the success of open source is that customer lock-in is both unnecessary and undesirable in terms of attracting and retaining paying customers. Being open lowers barriers to entry, and while it also lowers barriers to exit, that focuses the attention on to services (both in terms of meeting support service level agreements and providing value added application services) as a means of encouraging customer loyalty.

That encourages open standards and APIs, but not necessarily open source, although I do believe that the potential for open source to provide an on-ramp to the cloud could encourage open source as well.

One of the reasons for that is based on the evolving business strategies that are being applied to generate revenue from open source software, a subject covered in May.

Commercial open source vendors targeting on premise deployments have previously made the open source software available to community users while targeting paying customers with a combination of potential value added services, including support, custom development, subscription services, commercial extensions, or a proprietary-licensed version.

Choosing which services to choose and which features to make available to the community or customers is potentially complicated. Offering the software as a service (whether hosted, SaaS or cloud) potentially simplifies the issue by making an open source version available to the community and wrapping the same code in billing, management and provisioning capabilities for deployment as a service.

For the user the choice is simple: adopt the open source version to run in-house and support it yourself, or pay to run it as a service. With the emergence of private and hybrid clouds there is a third option for vendors, which is to offer support services for private cloud deployments and/or to assist in migration of applications and data to and from external clouds.

Cloud platform providers have so far been reticent about opening up the code for their platforms. With open source potentially acting as an on-ramp to the cloud we believe that the emerging cloud challengers will see a benefit in releasing their code as open source to encourage on-premise adoption and ease migration to public clouds.

This could, of course, be done with closed source software, but the benefits of open source in creating de facto standards and an ecosystem of developers weigh in favor of open source.