GNOME hacker Christian Schaller posed the question What do the free [desktops] need to grow in market share?. The comments are instructive.
It's easy to identify what individual commenters care about (and how they use free desktops and other software). The biggest thing the free desktop needs is either or all of:
- More proprietary software
- Killer applications for media creation and editing
- Easier deployment tools for enterprises
- Fewer bugs
- More handholding for home users
- A better OpenOffice.org and other migration-minded applications
- A killer application of any kind
- A shift to light-use Internet-based computers or devices
Strangely, I didn't see "device drivers" on that list -- nor did I see "marketing" or "pre-installation".
Though I have tremendous gratitude for everyone who's contributed code, money, attention, documentation, bug reports, and everything else to building a free operating system which does everything I need to do (and has done so for me for the past decade), perhaps the command-line developer and sysadmin community has its needs met adequately for now. That is, the free desktop provides sufficient tools and support for programmers and system administrators to do their work effectively. Now it's time to look elsewhere for growth.
Does that mean that freedom isn't the killer application for a free desktop? It is for me, but I didn't expect to see "people aren't demanding free devices and free software" either. Even so, the instructive examples of Google Video rendering purchased videos useless by closing, Wal-Mart briefly leaving its DRM servers running, and Xbox 360's restrictive download DRM provide a better example to non-techie users of why giving someone else control over what you can do with your computer isn't always pragmatic than twenty years of essays on why freely-redistributable source code is important. Freely-redistributable source code is important, but its appeal is mainly to programmers and other people willing to apply patches and tweak things as necessary.
The most striking thing about the comments isn't what they want or need, but the realization that there are plenty of opportunities to build software and systems orders of magnitude better for some system of measurement which can attract users from proprietary and unfree systems. I don't believe price can be an order of magnitude better (at least while Mac OS X is a differentiator for a slim line of hardware platforms, and while the effective cost of MS Windows is effectively zero for bundled deals). I'm not sure freedom on its own is an order of magnitude better, at least not on its own and for its own sake.
Some people might despair that all of the other "just one more thing!" requests are self-contradictory, but in my mind that's a strength. I'm happy to work on building the next great programming language while someone else builds a great audio editor and someone else writes the best educational software available. The freedom to combine and redistribute and modify tens of thousands of individual projects to meet any and all needs for a free desktop targeted at almost any field of endeavor makes me believe that the success of free desktops (or mobile devices, or embedded devices, or any general purpose computing device) can be inevitable. Even if the single necessary application or combination of features for any particular audience is too expensive or time-consuming to build or if the audience is too small for any company to care about making and selling a product, free software has the opportunity to bypass the old economics of proprietary software.
We don't all have to work on the same project or care about the same audience, as long as we care sufficiently about the freedom of those potential users to allow free and unfettered use, modification, and redistribution of that work. Freedom may not be the killer application of the free desktop, but freedom is the free desktop's sine qua non.