Earlier today, users reported widespread problems while trying to access Apple's cloud-based services, such as the iTunes Store and the Mac App Store. Although I didn't run into any difficulties outside my normal experience - so far today, I've downloaded Xcode for one of the MacBook Pros in the lab, so that my boss-dude-person could mess around with Installer packages and passwords, as well as updates to a couple of apps that live on the iPhone I use for testing - it's not a good sign, given how invested the company is in its forthcoming iCloud.
The iCloud service will be, along with iOS 5 and Mac OS X "Lion" (10.7 to those of us who keep track of such things), one of the main draws at Apple's Worldwide Developers Conference, which starts Monday in San Francisco. (Coincidentally, Monday is also the 67th anniversary of "D-Day" for the Normandy invasion of World War II. I wonder which blogger will be the first to use "storm the beaches" or a similar cliche.)
Speaking of WWDC, Apple is said to have released a app for that today, with scheduling features and maps of Moscone West. I use the expression "said to have released" because a search of the App Store around 1pm Pacific Time - and another at 3pm -turned up nothing but a 99 cent countdown clock. Not that it really matters to me; being a San Francisco resident, I can find Moscone West with my eyes closed, and being a member of the press, I'm not really welcome at WWDC. Apple graciously allows us media folk access to the keynote address - and nothing else. But since it's Apple's clambake, and that policy's been in place for years, I don't take it personally.
But back to my main point: despite the talk of how inevitable the cloud is, I will continue to be skeptical of the pie-in-the-sky claims being made on its behalf. Sure, the cloud can be convenient, and can keep TCO to a minimum. But if you're dealing with a service that guarantees 15 minutes or less per month of downtime during normal business hours, if those 15 minutes happen to coincide with your boss demanding a report from a cloud-based service right now, someone's going to lose.
There are some places where the cloud makes a great deal of sense, and backup is one of those. Although I'm not sure I want to try a full system restore from a cloud-based service such as Dolly Drive, such services are a very good alternative to maintaining a local backup and shuffling drives from one location to another. But in the event of the kind of disaster that would render a local backup unusable, a full system restore is exactly what I'm going to need to run, and even on the lab's relatively speedy (20Mb/sec) WAN connection, that's going to take a few days. I shudder to think how long it might take to restore even a subset of my data over anything slower than a T1.
So count me in the group of people who are excited by the potential of cloud-based services, but who don't believe that there's enough of the expertise needed to make these them work properly and consistently.