It dawned on me during a discussion on platform-as-a-service (PaaS), the latest software industry buzzword, that I had heard it all before. The idea that developers can build applications using a "platform" that deploys the application and its services in an on-demand model was mighty familiar.

It was about 3,000 years ago in techno-time that I first heard of the concepts behind PaaS, as well as business process modeling, enterprise service architectures, and other technologies now vying for your attention and budget. In essence, the principle is this: Separate the business use of software from development and deployment so that the user's requirements can be met without developers having to worry about how the final application runs in the real world of hardware and software.

This was the impetus behind the shift from mainframe to client/server architectures, and it has been a fundamental goal of technology companies ever since.

In the mainframe world, business logic, systems services, and hardware were all inextricably linked, to the financial benefit of IBM — the mother of all that was monolithic — and to users' detriment. Client/server was designed to break that mold by letting business functionality reign supreme while hiding the messiness of technology, and almost every important technology breakthrough since — including PaaS and its kissing cousins, cloud computing and on-demand — has followed that lead.

We know that client/server only partially succeeded, hence the ongoing search for the next big thing. What this tells us about PaaS and cloud computing is that we are destined to struggle with this dichotomy between business and technology for some time to come.

Don't get me wrong, I like PaaS and cloud computing, on-demand, business process modeling, and services-oriented architecture (SOA) because every one of them marks a significant advance toward solving this business use vs. deployment problem. But because these concepts are advancing simultaneously toward the same goal, we will likely take a few steps back as we move forward.

Whenever a new concept arises, its proponents try to leapfrog the previous panacea — as well as the state of the art inside real-world IT shops — ignoring the need for continuity. Thus, cloud computing and PaaS don't connect up with business process modeling: Most BPM solutions can't target the cloud, while most enterprise service architectures can't target the mix of cloud and on-premise computing that clearly will become the norm as the industry evolves. And many PaaS solutions require a new applications development environment that doesn't fully leverage what is already in use.

This continual change keeps organizations — manufacturers included — on their toes, but it requires level-headedness about when to move to the latest technology. All of this progress is supposed to be about supporting business efficiency and change, so make sure your business processes and strategies are in order first. Otherwise, you may find that in chasing the latest technology, you're chasing your own tail, trying to map imperfect processes to a new technology paradigm.

In the end, the more things change, the more they stay the same. And sometimes standing still is the best way to move forward, despite the promises inherent in the next new thing.