rms, Mr. GNU Project himself, recently interviewed with the Guardian and came out against "cloud computing." While there wasn't a great "splash" on the blogs I read--no one was surprised, a lot of people disagreed respectfully, and we all went along our ways--but I think he raises an important point that we/I should discuss.
rms' argument is that "cloud computing" services, like gmail, twitter, livejournal lock people into their services and force computer users to give up control of their computing and data, and that this is as bad or worse than using proprietary software.
I too have been a pretty big opponent of some "cloud computing" developments, not simply because they restrict freedom in the sense that rms is speaking of, but also because in a lot of cases, it's not a very good or user friendly environment for a lot of tasks. And I'm pretty stubborn about not trusting my data to a format that I'm not positive I'll be able to export into an open and useable format.
As it turns out there are a number of cloud services which are "more open," than others. That allow users lots of very standard/open access to their data. It's not the same as open source, of course, but I think it's important to suggest that all "cloud" services aren't created equal. Here's a brief review:
1. GMail provides full pop and IMAP access to their accounts. They also make it easy for you to use your gmail account to send "from" other accounts. While you get more freedom from hosting your email on a server you control, gmail isn't less free (in my estimation) than using the email address supplied by your ISP or employer. 2. Gcal similarly exports to ical format, which is the default calendar standard format, among others. 3. Most "cloud" news readers export OMPL files of all the feeds, similarly the "make RSS feeds of everything," mentality of web 2.0 means that a lot of data is pretty open to access.
There are of course some web-apps where data is opaque and not easily/openly exportable. Twitter is a great example of this, and I'm blanking on a more extensive list, but I'm sure that you, dear readers, can come up with many more examples.
I think what this means in the end is that the problem of freedom in software in the age of service-based computing rests more on user education than on a legalistic mechanisms (like copyleft) or open source code. Not that the later isn't important, but I think it's worth considering.