DesktopX - half of Stardock's attempt to recreate Microsoft's lost Cairo project - was publically unveiled today in alpha form for widespread download.
Ever it since it was burned by the failure of the Bob project, Microsoft has been cautious about throwing new UI paradigms at the great unwashed. DesktopX not only does this - with a vengeance - but smuggles in the first end-user scriptable object framework we've seen since OpenDoc. It's very gracefully subversive, and has the potential to do for Windows what Hypercard did for the Mac in the eighties.
What's slightly surprising since we last looked is the quantity of useful objects that are available. DesktopX objects don't just look purty, there's one that controls WinAmp, another that's a visual biff for checking your mail, and another that's a replacement drive icon that shows free disk space as a pie chart.
Most of this has been achieved without programming - all the user has to do to create their own is to call up a template and start filling in the blanks from dialog boxes. A browser is not far behind, says Stardock CEO Brad Wardell. "That should be possible in a few hundred lines of code," he claims. It's still early days and the objects haven't really begun to interact with each other.
So what doesn't it do? Well it doesn't fix any of Windows' underlying problems - terrible hardware latencies for multimedia, registry bloat, security ... we suspect you know these pretty well. But that would hardly be fair, as it doesn't claim to. It isn't cross platform either, and there are no plans to implement it on non-Windows operating systems.
However it's a idea bigger than any one platform, and extremely well implemented, and so Stardock could benefit from tolerating the kind of benevolent reverse engineering that AOL and Napster demonstrated with ICQ and Napster the client.
Of the many impressive ideas floating around Linux GUI land, none makes the break with the Mac/Windows metaphor, although as a development exercise, writing cross-platform DesktopX objects would be far from trivial.
A few rough edges remain too, and on release could benefit from some wizards, and very clear "what do I do now?" signposts. Corporates will almost certainly want the ability to turn disable many of the add/modify object functionality - they're too damn easy to get to (which of course, is the point).
As we noted when we first heard of the project, it could prove to have a major strategic influence on future Windows, as OEMs have been dying to take the "user experience" into their own hands for some time, and one of the proposed DOJ remedies put limits on Microsoft being able to prevent this. But as the Microsoft trial heads for the appelate court, with every prospect of a mini trial rerun stretching well into next year, that day won't come anytime soon.
It's version 0.50, although Stardock.net subscribers have been treated to early previews. Windows 2000 users get a revamped task list, and alpha-blended objects. Wardell doesn't expect a release before the end of Q2 next year. Stardock plans to rewrite configuration files in XML, so if the Windows goalposts move, it'll be easier to move DesktopX with them.