Now you're just playing me, 40hz~! 
You somehow knew that I have a very soft spot in my heart for Aesop! 
(Among my favourite books is "Aesop without Morals", which is an original translation. Great read! ) Playing on getting my sympathy there... consider it done! -Renegade
Perish the thought I'd stoop to something like that!
The key there is "linked".
We all know darn well that application software doesn't require a "linked" environment to run. So why pay for that? It's just milking people. Nothing more.
Now, there is value in the "link", but that's in the data transmission, and not in the application.
Cloud storage is a reality now.
The keyword is "linked" in the sense of sharing enabled. Which slides in neatly with shared storage. And "ubiquitous access" as in "my stuff anywhere and on any machine."
Do you absolutely need it? Depends. But it's nice to have. (And that's
not dependent on Microsoft as we both know.)
But if you can
incorporate it into an already popular product line and make it available with a simple mouseclick, it's just one more step towards locking your customers in.
Do you absolutely need it. Depends on what you do. But I found Skydrive to be a blessing when moving docs back and forth between me and my clients. Sure beats e-mailing files back and forth. And all without the need to either set a client up on my server - or me on theirs. Something that's becoming an increasingly touchy subject for many of my clients with all the concerns about network security these days.
So for moderate, ad hoc, or short-term sharing, it works very well.
Cloud computing is not. And won't be for a while. It's coming, but not here yet.
We simply cannot run heavy duty applications in the cloud. The network just doesn't support it. Yet.
Yes and no. For the average home broadband service on DSL it can be touch and go. But it works much better than I thought it would. On a cable or other standard high-speed connection it's remarkably good. On a dedicated corporate-level fiber connection it just works. Period.
The thing with cloud computing is that to really tap the benefit it will take some rethinking and recoding of many apps to take them from a dedicated PC/CPU architecture to a client/server design. Nothing insurmountable. There's plenty of precedent and best practice behind doing that sort of application. But even so, it's seldom necessary or desirable to have the actual app running on the host unless it's doing something like a huge rendering or gene sequencing task which needed a cluster in order to finish within an average human lifetime.
But it doesn't need to be
all cloud or
all local. Some sort of PC will always need to be around so you can also do a hybrid approach. On the client side all you'd need is a client app (which
would do the heavy lifting in most cases) and something to link to the backend. Most likely for little other than to obtain an authorization token to prove you are currently an active subscriber. That and to take advantage of the increasingly ubiquitous online storage being offered with many apps.
In this scenario, the app is
installable via the cloud and is
local on your machine. You only need to be able to connect to the web at least once every thirty days for it to greenlight you. Think of it as a web-based version of the old antipiracy hardware dongle all those expensive graphics and music software publishers used to be so fond of.
Adobe is already doing it this way - and I personally think their model is the one that's going to win out long term.
I'll happily side with the wolf there. 
Yeah. Me too.
