Nortel Networks’ announcement last month of an Open IP Environment that could run on a variety of devices regardless of who the devices’ vendor is seems like a great idea. If vendors can’t rely on proprietary software to lock users into implementing only one specific vendors’ equipment, they can’t charge inflated prices. Unfortunately, despite the potential merit of Nortel’s plan, its eventual success is very much in doubt.

At a glance, Nortel’s strategy appears infallible. What network manager wouldn’t kill for an entirely open, vendor-neutral network where one application would interact with other applications and hardware devices seamlessly? If such an environment existed, network managers would be able to buy the most technically sound equipment available at the best price.

What’s more, Nortel’s Open IP Environment would distribute routing functionality, reducing the existing bottlenecks at access routers.

Networking hardware vendors today do a good job of talking standards, but when it comes to implementing features like QoS, each vendor does so in a different way. What works with one vendor’s boxes probably won’t work with another’s.

So what’s wrong with Nortel’s plan for an open IP environment? One drawback is that it’s still just that — a plan.

Nortel has signed an agreement with Intel, claims 75 other development partners and is including elements of its open IP software in Microsoft’s existing NT platform and future Windows 2000 operating system. But these are just baby steps. There’s not telling how long it will take Nortel and any application developers who sign onto the open IP initiative to get mission-critical, well-tested applications out the door.

The second, and probably bigger, problem with Nortel’s plan is that in order to succeed it has to overcome Cisco Systems Inc. and its massive installed base of customers.

By almost all accounts Cisco owns around 70 per cent of the worldwide router market share.

While many observers would label Cisco a hardware company – after all it does manufacture hardware components such as routers and switches – as one anlayst recently noted in a conversation with me, it could just as easily be labelled a software firm.

Cisco’s hardware isn’t anything very special on its own. Cisco boxes probably won’t win in any “feeds and speeds” face-offs against competitors like Foundry Networks. What differentiates Cisco equipment from everything else on the market is the software that runs on it and the support Cisco throws behind its gear.

With the latest release of its Internetwork Operating System (IOS), Cisco says it is offering users the ability to prioritize applications by reading into IP packets. The catch is users must be using Cisco equipment on both ends of the connection they want to apply application priority to.

Part of the goal of Nortel’s open IP initiative seems to be to remove Cisco’s edge in the software category. If developers flock to Nortel’s environment and create applications that will run in any networking environment, Cisco users will no longer be forced to buy Cisco gear to get some of the value-added software features they have come to enjoy by running Cisco-only networks.

Over the years though, Cisco’s customers have proven to be fiercely loyal. The company’s router market share appears to be holding, despite the fact that competitors are offering users cheaper gear. Obviously, Cisco’s customers are happy with their networks’ performance and the service and support they receive. What makes Nortel believe this will change in the future?