If Linux and open-source developers in general want a good idea of a project to take cues from, there's one from Microsoft that is worth a long, hard look. No, not Vista -- in fact, it's not a desktop product at all, strictly speaking. It's Windows Home Server.

Windows Home Server ("WHS") has been earning praise from people who normally have few kind things to say about Windows or Microsoft in general. WHS is sold either in conjunction with a headless server (Hewlett-Packard's MediaSmart server is one such device) or as a standalone software product you can add to hardware of your choice.  It provides a whole slew of services for up to ten Windows desktop machines: media and file sharing, remote access, printer management, and total system backup with bare-metal restore. The backup system, my favorite feature so far, requires zero intervention on the user's part and runs on the cluster and not the file level, so only changes to files are written across the wire. In short, WHS is turning a lot of heads -- mine included.

Why am I bringing this up as an example of the kind of project that FOSS developers might want to pay attention to? For one, it's the sort of thing a good Linux team would seem perfectly suited to creating -- something they could use to draw attention to Linux in a great way.

Linux already has just about all the server components to build something like this, ready to go.  What they would need to do on top of that is write at least two other things for it -- a good management interface, and a Windows client to handle things like the seamless backup functionality.  If they could offer something like WHS at lower cost to the end user, or at the same cost but with an even better spate of features, that would be a real head-turner.

If you ask me, one of the best ways Linux can try to make inroads to the desktop (and probably a great many other places besides) is by more explicitly complementing Windows instead of trying to outright eclipse it. WHS seems like a good model for one such project.