Cloud computing offers many advantages over today's siloed infrastructure, but let's face it -- getting from here to there won't be easy.

It's not so much technical hurdles that stand in the way, but more the fear that something new will lead to unforeseen complications, placing valuable data at risk.

That's to be expected, say proponents of cloud systems, which is why most organizations are test-driving the technology on non-essential stuff before entrusting it to mission-critical operations.

According to speakers at the Structure 08 conference in San Francisco this week, the major concerns over clouds center on security and control. How can data held in third-party, on-line storage systems be kept safe, and how can organizations maintain control of their data once it leaves in-house resources?

The answer to the first question is easy: it can't. But it doesn't matter whether your data is in the cloud or the local array, data can never be 100 percent safe, which is why the data security industry has mushroomed over the past three decades. The answer to the second question is a little less clear, although it's reasonable to assume that data management will be a key area of development as cloud technology evolves.

One of the key things that cloud boosters need to understand, though, is that it's not up to the enterprise to wake up to the wisdom of clouds. It's cloud that needs to make a compelling argument to the enterprise. Google's Christophe Bisciglia didn't seem to get this basic fact at Structure 08, likening the fear of clouds to the fear of banks in the old days. Attitude adjustment is fine, but it can only come about if those providing the cloud take concrete steps to allay the very legitimate concerns of CIOs.

There's also the very real possibility that certain applications and data sets will never be ready for the cloud. As Dave Rosenberg, CEO of open source developer MuleSource, points out, while back-office and business intelligence systems might do well in the cloud, software governing things like bank transactions, stock trading and many telecommunications functions won't.

Most of these arguments, though, revolve around the concept of the cloud as a Web-based service. But as we've pointed out before, there's no reason why organizations can't set up cloud environments behind their own firewalls. Although probably more expensive than the cloud-as-service model, it alleviates many of the worries over security and control. As this anonymous blog on Appistry points out, it was the Internet itself that led to the rise of corporate intranets, so the idea of a universal technology being adapted to private use is not without precedent.

For the cloud to be a successful enterprise tool, it's going to have to develop according to the enterprise's terms. As the newcomer on the block, it's simply not realistic to have the world jump through all the hoops.