I just read an article in New Scientist about cloud-based music storage & streaming sites, asking why anyone would want to own music, if it was all available in the network. I've heard the same argument advanced about image or video or work data storage.
Don't get me wrong, I think Spotify is wonderful and I used Rhapsody for a while. I stream Internet radio to my phone over 3G. And I enjoy numerous other "cloud" services such as webmail and Facebook.
But for all the greatness of those services, they also highlight various limitations which mean that I absolutely want all my music and photos and contacts in "owned" form, on a hard drive or flash memory or CD or whatever.
Having the cloud as a backup is fine - or as a convenient way to get access to the "long tail" of things I'll never buy, yes. But as a primary source and repository for my most important mobile stuff? No thanks.
(For fixed applications, things are a bit different - this argument is specifically about the 'mobile cloud')
So what are the limitations?
The first one (as ever) is mobile coverage. It's still far from ubiquitous, even in central London. Obviously the Tube still doesn't have mobile coverage. But even just normal 3G at street level, or indoors, is patchy. Yesterday afternoon I was at an office in Holborn in central London. No 3G on Vodafone, approximately 100m from global mobile HQ - the GSMA is based a block down. Yes, we'll see improvements with femtocells and 800MHz spectrum and assorted other innovations - but I seriously doubt we'll ever be happy enough with coverage to trust it for data or content we really need now.
The next is capacity, especially for uncompressed data. If I take a photo with an 8MP camera, I'd like to retain a version that's comparatively unmangled - the ability to generate data volumes is accelerating beyond the uplink capacity of mobile networks, as is the propensity of people to generate content such as video. Yes it could be cached and uploaded later, but the sheer volumes make it sensible to store most of it locally, especially as memory is so cheap.
Then there's roaming. The idea that anyone wants to sit on a beach on holiday, pulling down music from the "cloud" at prices that insult the intelligence, is not believable. In fact, the whole data roaming scenario is now such an abuse of trust, that it will act as a barrier to mobile operators succeeding in new marketplaces. Honestly - do they think anyone will be stupid enough to trust an operator with their content - or worse, bank account - if they continue to perpetuate daylight robbery when you go across borders? It sends out a strong signal that "loyalty" is not wanted, just gullibility and compliance.
Linked to that issue is churn. Yes, I know that many services are intended to be "sticky" and reduce churn. But people do churn, for various reasons, and will continue to do so. Blackmailing them with potential loss of their data or content is unlikely to be seen positively. The main obstacle here is anything that ties service to access - that is clearly a link that will be broken even more fully than it is today.
Lastly - longevity. Inevitably, some "cloud" providers will go bust at some point. What happens to your data?
(There's also plenty of other issues around battery life, security, privacy and so forth).
Bottom line - mobile cloud services will indeed become more important and valuable. But the notion that they will replace ownership (and local storage) of data and content is risible - wishful thinking by those that want perpetual services revenue streams.
People like to own stuff - not rent everything.
Or else perhaps you'd like to sign up for my new mobile footwear service? A free pair of shoes on a 24 month $25 contract. The bundle includes 25km of walking a month, with an optional $3 a month upgrade for football-kicking. Unless you're travelling abroad, in which case you have to pay an extra $20 every time you want to wear sandals on the beach.