Cloud computing faces challenges as a straight datacenter-outsourcing model, but the cloud is clearly the only way to go for delivering mobility applications.

Craig Mathias has a column at our sister site Information Week, where I think he homes in on the obvious place where cloud architectures will succeed--namely, in serving mobile users. He writes:

Something profound is happening, Apart from games and a few other apps, most of those apps are really front-ends for applications which are essentially Web services, or otherwise running in the cloud. The real issue is formatting for the small screen, not local execution. So apps aren't really as bound to devices as is the case with the PC.

That's what the info-centric era is really all about -- moving IT resources into the cloud, and enabling access for authorized users on a very broad range of devices, from PCs to a wide variety of otherwise incompatible handsets. The computer no longer matters -- only the information does. (his emphasis)

Cloud skeptics like Melanie Turek have discussed the challenges that cloud computing faces as a straight datacenter-outsourcing model, especially for communications applications, but the cloud is clearly the only way to go for delivering mobility applications.

The challenge for IT is going to be managing this mobility. In his Info Week column, Craig also writes that we'll see users "selecting (yes, they'll own them) the most appropriate device for the job with no additional requirements placed on IT. I can even see borrowing someone's handset, using smartcard-based two-factor authentication, doing what I need to do, and then handing the device back with no concern about security" (my emphasis).

There are people I've met at VoiceCon from the enterprise world whose blood would run cold if they read that--users selecting and buying their own devices, using them willy-nilly and sharing them like so many sticks of Doublemint (in the IT manager's imagination). Sure, Craig imagines users following appropriate security measures, and many will, but....

Furthermore, there's the cost issue. Right now, it's hard to imagine this mobile connectivity to the cloud happening over anything but public cellular networks--probably the least cost-efficient way in existence for the enterprise to connect its users to their services.

Nevertheless, Craig is probably right in the basic picture he paints: Users will drive the next generation of mobile-based work, they'll do it in the way that is most convenient and efficient for them, and IT will be in the business of keeping up with them.

Maybe that's the new model for IT: Not as the organization within the enterprise that establishes and doles out connectivity and computing devices to the user population, but rather as the organization that manages the connectivity and computing devices that users provision for themselves.

I said "manages." Maybe that should be "struggles to manage."Cloud computing faces challenges as a straight datacenter-outsourcing model, but the cloud is clearly the only way to go for delivering mobility applications.