At the risk of truly reducing my perspective here to the absurd, the reason we buy computers of any form is to run applications. It therefore stands that the reason we buy networks -- wired and wireless -- is to run applications. The whole point of IT is, in fact, to run applications and manage the data they use -- locally, on servers, across enterprise networks and on the Internet.
And the whole point of contemporary wireless networks is to minimize the behavioral and performance differences that separate wireless and wireline, thus enabling applications to run with ease across both domains. We've largely accomplished this -- 150 Mbps on 802.11n, on the order of a megabit on cellular (with much more on the horizon), the commonality of IP, etc.
In fact, the only key difference between the two that exists today, apart from a greater degree of throughput variability on wireless, is the limited availability of wireless. Sometimes it works just fine, sometimes it s a little flaky and sometimes, well, it s just not there at all. And arbitrary transitions between these modes remain all too common.
This challenge will continue to be addressed and will eventually be solved in 3.5G, 4G, and especially converged cellular/Wi-Fi networks. It has to be, because the core justification for mobility is the improvement in productivity that results from anytime, anywhere voice and data communications capability. It's simple: no connectivity, no productivity.
One other issue: the wide diversity of application platforms on smartphones, which makes software development, um, interesting.
Add all this up, and it's clear we're not yet at the point where applications can move freely between the wired and wireless domains. But there is significant progress to report in moving towards this goal:
Cisco's recent announcement of its new 3300 Series Mobility Services Engine recreates a concept known as mobile middleware as a network service. Mobile middleware has been around for a long time. It was initially conceived as software libraries that would decouple applications from (and delay the binding to) wireless networks, and to make it easy to port or develop apps on handsets and similar devices. Unfortunately, there are more than 200 vendors of proprietary middleware products, and a good number of them weren't really very usable (or even useful) at all. But the concept is sound, and Cisco has now put a broad range of extensible cross-network mobility services on a new class of network server. This is an important direction for mobility and will spur the development of a broad range of enterprise apps and services.
Mobile middleware itself is not dead. Check out Vaultus Mobile Application Platform, which makes it easy to map applications running on a server onto a handset. An alternative approach can be seen in Cascada Mobile's soon-to-be-announced Breeze, which allows anyone with a knowledge of HTML to easily build business-oriented apps for a variety of handsets.
The iPhone 3G and Apple's developer kit for it show what can happen when a handset manufacturer really commits to mobile applications. I expect hundreds, if not thousands, of iPhone apps over the next few years. Sure, most of these will be consumer-grade at best, but some will help to propel the iPhone to contender status as a serious business tool. And it would be foolish to discount the impact of LINUX on future handsets, most visibly today in the form of (but not limited to) Google's Android effort.
I expect that once we solve the variable-connectivity problem I noted above, mobile applications will focus on the Web-services model, easing software development. In reality, a hybrid approach will almost always be required, with the client/server model prevailing because it simply doesn't make sense (nor will it be even possible in most cases) to move all of the data required by an app to a handset. Regardless, as I noted above, enterprises that plan for a mobile IT future will be building the flexible and functional infrastructures that will boost productivity and help to propel their businesses going forward.
Craig J. Mathias is a principal at Farpoint Group , an advisory firm specializing in wireless networking and mobile computing. He can be reached at firstname.lastname@example.org.