I sometimes don't see the use of outsourceing basic stuff from a company, just because it might be cheaper. For computing, I like distributing work to all the company's PCs, workstations and some servers better than sending it off somewhere. That would be a more desktop-based approach, like BOINC

Using the allready existing resources better is more important than pushing another centralized outsourceing solution forward.

I see "cloud" computing and think "mainframe" and "dumb terminals" and "thin clients".

All are good things form a certain point of view and one I support to a degree.

I'm OK with cloud computing but I'd rather have my data backed up where I can see it.  And when the cloud isn't there, I still need to work.

I think the solutions will attract a very specific kind of customer, and it won't be gamers. I myself frequently work on a remote desktop set up in my home. Togheter with mobile broadband, this could be the start of a new era regarding how computers are seen. Or it could be a minor footnote. I think which way it goes will fall on how well the applications produced for this specific need are made, to fit a target audience with relatively low demands. Ultimately, and I know some will disagree with this, if they can produce competitive complete products (~full systems) at a lower end-price for the consumer than the current options, the sky is the limit.

Does anyone remember that movie where they produce a flashlight and it's a whole computer? That's what I am thinking this is a step towards (specifically, two of the features of the flashlight were storage and processing "online").

I think lightning, storms and colliding weather systems.

Didn't Microsoft recently register a patent for something like this? Will the computers start to devolve back to the way they used to be, as Zubaz mentioned? As an about to graduate professional, i find this deveploment very interesting.

Yeah, I see this as well.

And in all honesty - what's good for businesses isn't always good for us home users.  Frankly, the 100% internet is a bit of a dream, and there are still security, privacy, latency, bandwidth, trust, etc etc etc issues.

Frankly, I don't think it's even a philosophically good idea to go 100% cloud computing.  There's nothing inherently evil or wrong about offline or hybrid applications.  I just don't get why people are so driven to make this happen to everything under the sun.

I can understand it for some applications.  But everything?  Really?  Why?  And why 100% "eventually?"  Why not eventually hybrid?  What's so horrific about offline computing that we have to think that 100 or 1000 years from now we'll all be "100% on the cloud?"

Frankly, I think it's more dogma than sense.

My opinion is that 100-1000 years from now, we're going to be using systems that will seamlessly work both online and offline.  Hybrid systems with seamless synchronization.  That's my vision, and IMHO it makes a lot more sense.

And franky, some things may never be totally or even partially online, even 1000 years from now.  Sensitive medical info, business trade secrets, and classified information are things that will always make sense to remain isolated from the general Internet.