Novell CEO Zigs At Interop. Meanwhile, Web 2.0 (Across The Hall) ZagsNovell CEO Zigs At Interop. Meanwhile, Web 2.0 (Across The Hall) Zags

If you're not here at the Jacob Javits Convention Center in New York City, then what you may not know is that Interop and Web 2.0 Expo (both sisters to information under the TechWeb umbrella) are running side-by-side with one another in the same building. What's really odd about this is that if you're at Interop, you can almost feel the Web 2.0 folks across the way disruptively changing the course of IT history while those here at Interop continue to evangelize the tried and true approach to IT. As evidenced by Novell CEO Ron Hovsepian's keynote, history is clearly repeating itself.Hovsepian (pictured below) practically harkened back to the early days of Interop when the event was primarily about interoperability. The major them of his presentation (and clearly Novell's strategy going forward) was "making IT work as one" and one of the key pillars to that strategy -- a pillar that Hovsepian says Novell is positioned to help with -- is interoperability. Hovsepian talked about how, IT organizations should be leveraging interoperability and standards to seamlessly blend infrastructures at various layers in the IT stack. For example, virtual systems and physical systems.

But with Web 2.0 happening across the way, a real "say what?" moment (at least for me) happened when he told the audience that the focus of their development efforts should be on Microsoft's .NET and the enterprise version of Java (he called it by its old name "J2EE" but today, it's really just called "JEE"). At the application development level according to Hovsepian, these are also infrastructures that must interoperate so as to "make IT work as one."

Perhaps it was sage advice for the IT audience in attendance, which is clearly as conservative as it has ever been judging by the show of hands that were raised when, at one point during the morning's keynote round, they were asked how many of them had moved some of their IT into the cloud. Hardly any hands went up. But, if you ask me, history is just repeating itself.

Almost two decades ago, as the client-server revolution was starting to pick up some steam, there was no shortage of players in the market that were still advocating the older, but admittedly more proven approaches to application infrastructure (including continued investment in terminal-driven mainframe applications).

Over time, the industry has experienced several revolutions. In the early 80's, it was PCs. Then came local area networks, client-serving computing, Internet and the Web, services-oriented architectures, virtualization, and more recently cloud computing (your interpretation of computing history may be a bit different). Many of the organizations that embraced the new approaches of the time were able to cash in on their promises. It didn't always work out though.

Today, the new approach is very much about the cloud and the deconstruction of functionality into more discreet and distributed software components. .NET and J2EE have their advantages. But in the new regime, you have wholly programmable platforms (aka: Platforms As A Service) in the cloud like salesforce.com's Force platform, Google's Google App Engine and Bungee Labs' Bungee Connect that, to J2EE and .NET are exactly what each of the aforementioned revolutions represented to their predecessors (or that which they were disrupting).

Across the course of IT history, there has been no shortage of IT shops that stuck (or are still sticking) to platforms that have reached their golden years (not that I see JEE or .NET as being ready for the pasture any time soon). So, that part of what Hovsepian had to say undoubtedly resonated with a great many people in the room.

But what struck me as being odd is that I heard no acknowledgment of the newer approaches which, quite frankly, for most organizations will eventually end up as a part of the mix that Novell aims to "make work as one." For example, in the course of seamlessly blending physical and virtual infrastructures, perhaps some number of virtual machines running in Amazon's Elastic Compute Cloud will have to be drawn into the formula as might applications running on salesforce.com's Force platform or Google's App Engine. What about scripting your way to application success using PHP or Python (the latter of which is supported by Google's App Engine). In Hovsepian's favor, Python can also be used to script .NET and JEE.

But something tells me that that's not the .NET and JEE he was talking about. Not only that, if you head over to Web 2.0 land across the way, finding applications driven by Python on Java or .NET won't be easy.

Or how about Ajax as the compute engine (yes, in the browser) drawing upon Internet-bound functionality (e.g., maps) from the cloud through application programming interfaces (APIs). In that case, the Internet is "the platform."

In a great many cases, Web 2.0 is driven by a good ole' LAMP stack (P for PHP and Python) that in some cases is spread across cloud services like Amazon's EC2 and/or Simple Storage Service (S3). Occasionally, Web 2.0 gets poked in the eye when the cloud "goes down." But Web 2.0 is not a flash in the pan. It is, in fact, a proof point that scalable and reliable systems can be built using what many in (and who speak to) the Interop audience would probably consider to be non-traditional if not non-mission critical or unorthodox approaches.

Over on the Web 2.0 side of the building where the new kids on the block are Twittering and Facebooking like mad from their iPhones, there may not be much in the way of solutions that enterprise IT people will find interesting. But make no mistake about it. The technology under the hood that makes Web 2.0 tick is no more or less mission critical than the what's found in any corporate data center. Sooner or later, corporate IT and the vendors that traditionally serve them will have to reckon with that very disruptive force.