Canonical calls it "Ubuntu One". The goal is, according to Stefano Forenza, to build a full-blown online platform -- Canonical's version of Windows Live, or MobileMe / .Mac / iTools. But can they do it without having their intended userbase siphoned away?

A detailed discussion of it is scheduled at OSCON, and the blurb there is quite telling:

This to me is a broad hint as to how Canonical plans to monetize Ubuntu: not just by selling support for it, but selling services -- end-user services that people actually use and want to pay for because they already do. If people are going to take their computing to go, and you want them to use your platform for same, best to give them both the local and remote ends of the deal. Give them both something to plug into, and something to plug in with.

Good idea. Here's what's worth worrying about, though: The development of the project is entirely internal, although the client software is GPLv3- / CC-licensed. So if the APIs and client software are all open, who's to say that they won't be ported to other platforms, thereby defeating the whole purpose?

Historically, that's what's happened with anything free/open. Random example: far more people run OpenOffice on Windows than they do on Linux (in fact, it's developed on Windows to begin with). Unless Canonical is willing to admit that they may ultimately be building a platform that is OS-agnostic, and are willing to live gracefully with that, it won't be worth the trouble.

InformationWeek has published an in-depth report on cloud storage. Download the report here (registration required).

Follow me and the rest of InformationWeek on Twitter.